O.dyer PID

PI and PID controller tuning rules for time delay processes: a summary
Technical Report AOD-00-01, Edition 1

A. ODwyer,
School of Control Systems and Electrical Engineering,
Dublin Institute of Technology, Kevin St., Dublin 8, Ireland.
15 May 2000
Phone: 353-1-4024875
Fax: 353-1-4024992
e-mail: aodwyer@dit.ie
Abstract: The ability of proportional integral (PI) and proportional integral derivative (PID) controllers to
compensate many practical industrial processes has led to their wide acceptance in industrial applications. The
requirement to choose either two or three controller parameters is perhaps most easily done using tuning rules. A
summary of tuning rules for the PI and PID control of single input, single output (SISO) processes with time
delay are provided in this report. Inevitably, this report is a work in progress and will be added to and extended
regularly.
Keywords: PI, PID, tuning rules, time delay.
1. Introduction
The ability of PI and PID controllers to compensate most practical industrial processes has led to their
wide acceptance in industrial applications. Koivo and Tanttu [1], for example, suggest that there are perhaps 510% of control loops that cannot be controlled by SISO PI or PID controllers; in particular, these controllers
perform well for processes with benign dynamics and modest performance requirements [2, 3]. It has been stated
that 98% of control loops in the pulp and paper industries are controlled by SISO PI controllers [4] and that, in
process control applications, more than 95% of the controllers are of PID type [3]. The PI or PID controller
implementation has been recommended for the control of processes of low to medium order, with small time
delays, when parameter setting must be done using tuning rules and when controller synthesis is performed
either once or more often [5]. However, Ender [6] states that, in his testing of thousands of control loops in
hundreds of plants, it has been found that more than 30% of installed controllers are operating in manual mode
and 65% of loops operating in automatic mode produce less variance in manual than in automatic (i.e. the
automatic controllers are poorly tuned); this is rather sobering, considering the wealth of information available in
the literature for determining controller parameters automatically. It is true that this information is scattered
throughout papers and books; the purpose of this paper is to bring together in summary form the tuning rules for
PI and PID controllers that have been developed to compensate SISO processes with time delay. Tuning rules for
the variations that have been proposed in the ideal PI and PID controller structure are included. Considerable
variations in the ideal PID controller structure, in particular, are encountered; these variations are explored in
more detail in Section 2.
2. PID controller structures
The ideal continuous time domain PID controller for a SISO process is expressed in the Laplace domain
as follows:
U( s) = G c (s) E (s)
(1)
with
G c (s) = Kc (1 +
1
+ Tds)
Ts
i
(2)
and with Kc = proportional gain, Ti = integral time constant and Td = derivative time constant. If Ti =
and Td = 0 (i.e. P control), then it is clear that the closed loop measured value, y, will always be less than the
desired value, r (for processes without an integrator term, as a positive error is necessary to keep the measured
value constant, and less than the desired value). The introduction of integral action facilitates the achievement of
equality between the measured value and the desired value, as a constant error produces an increasing controller
output. The introduction of derivative action means that changes in the desired value may be anticipated, and
thus an appropriate correction may be added prior to the actual change. Thus, in simplified terms, the PID
controller allows contributions from present controller inputs, past controller inputs and future controller inputs.
Many tuning rules have been defined for the ideal PI and PID structures. Tuning rules have also been
defined for other PI and PID structures, as detailed in Section 4.
3. Process modelling
Processes with time delay may be modelled in a variety of ways. The modelling strategy used will
influence the value of the model parameters, which will in turn affect the controller values determined from the
tuning rules. The modelling strategy used in association with each tuning rule, as described in the original
papers, is indicated in the tables. Of course, it is possible to use the tuning rules proposed by the authors with a
different modelling strategy than that proposed by the authors; applications where this occurs are not indicated
(to date). The modelling strategies are referenced as indicated. The full details of these modelling strategies are
provided in Appendix 2.
K e s m
A. First order lag plus delay (FOLPD) model ( G m ( s) = m
):
1 + sTm
Method 1: Parameters obtained using the tangent and point method (Ziegler and Nichols [8], Hazebroek and
Van den Waerden [9]); Appendix 2.
Method 2: Km , m assumed known; Tm estimated from the open loop step response (Wolfe [12]); Appendix
2.
Method 3: Parameters obtained using an alternative tangent and point method (Murrill [13]); Appendix 2.
Method 4: Parameters obtained using the method of moments (Astrom and Hagglund [3]); Appendix 2.
Method 5: Parameters obtained from the closed loop transient response to a step input under proportional
control (Sain and Ozgen [94]); Appendix 2.
Method 6: Km , Tm , m assumed known.
Method 7: Parameters obtained using a least squares method in the time domain (Cheng and Hung [95]);
Appendix 2.
Method 8: Parameters obtained in the frequency domain from the ultimate gain, phase and frequency
determined using a relay in series with the closed loop system in a master feedback loop. The
model gain is obtained by the ratio of the integrals (over one period) of the process output to the
controller output. The delay and time constant are obtained from the frequency domain data
(Hwang [160]).
Method 9: Parameters obtained from the closed loop transient response to a step input under proportional
control (Hwang [2]); Appendix 2.
Method 10: Parameters obtained from two points estimated on process frequency response using a relay and
a relay in series with a delay (Tan et al. [39]); Appendix 2.
Method 11: Tm and m are determined from the ultimate gain and period estimated using a relay in series
with the process in closed loop; Km assumed known (Hang and Cao [112]); Appendix 2.
Method 12: Parameters are estimated using a tangent and point method (Davydov et al. [31]); Appendix 2.
Method 13: Parameters estimated from the open loop step response and its first time derivative (Tsang and
Rad [109]); Appendix 2.
Method 14: Tm and m estimated from Ku , Tu determined using Ziegler-Nichols ultimate cycle method;
Km estimated from the process step response (Hang et al. [35]); Appendix 2.
Method 15: Tm and m estimated from Ku , Tu determined using a relay autotuning method; Km estimated
from the process step response (Hang et al. [35]); Appendix 2.
Method 16: G p ( j135 ) , 135 and Km are determined from an experiment using a relay in series with the
0
process in closed loop; estimates for Tm and m are subsequently calculated. (Voda and
Landau [40]); Appendix 2.
Method 17: Parameter estimates back-calculated from discrete time identification method (Ferretti et al.
[161]); Appendix 2.
* Method 18: Parameter estimates calculated from process reaction curve using numerical integration
procedures (Nishikawa et al. [162]).
* Method 19: Parameter estimates determined graphically from a known higher order process (McMillan [58]
also McMillan (1983), pp. 34-40.
* Method 20: Km estimated from the open loop step response. T90% and m estimated from the closed loop
step response under proportional control (Astrom and Hagglund [93]?)
Method 21: Parameters estimated from linear regression equations in the time domain (Bi et al. [46]);
Appendix 2.
Method 22: Tm and m estimated from relay autotuning method (Lee and Sung [163]); Km estimated
from the closed loop process step response under proportional control (Chun et al. [57]);
Appendix 2.
* Method 23: Parameters are estimated from a step response autotuning experiment Honeywell UDC 6000
controller (Astrom et al. [30]).
Method 24: Parameters are estimated from the closed loop step response when process is in series with a
PID controller (Morilla et al. [104a]); Appendix 2.
Method 25: m and Tm obtained from an open loop step test as follows: Tm = 1.4( t 67% t 33% ) ,
m = t 67% 1.1Tm . K m assumed known (Chen and Yang [23a]).

Method 26: m and Tm obtained from an open loop step test as follows: Tm = 1.245 ( t 70% t 33% ) ,
m = 1. 498 t 33% 0. 498 t 70% . K m assumed known (Miluse et al. [27b]).
* Method 27: Data at the ultimate period is deduced from an open loop impulse response (Pi-Mira et al.
[97a]).
B. Non-model specific
Method 1: Parameters K u , K m , u are estimated from data obtained using a relay in series with the process
in closed loop and from the process step response (Kristiansson and Lennartsson [157]).
need to check how the other methods define these parameters
C. Integral plus time delay (IPD) model ( G m ( s) =
K me s m
s
Method 1: m assumed known; Km determined from the slope at start of the open loop step response
(Ziegler and Nichols [8]); Appendix 2.
Method 2: Km , m assumed known.
Method 3: Parameters estimated from the ultimate gain and frequency values determined from an experiment
using a relay in series with the process in closed loop (Tyreus and Luyben [75]); Appendix 2.
Method 4: Parameters are estimated from the servo or regulator closed loop transient response, under PI
control (Rotach [77]); Appendix 2.
Method 5: Parameters are estimated from the servo closed loop transient response under proportional
control (Srividya and Chidambaram [80]); Appendix 2.
Method 6: K u and Tu are estimated from estimates of the ultimate and crossover frequencies. The ultimate
frequency estimate is obtained by placing an amplitude dependent gain in series with the
process in closed loop; the crossover frequency estimate is obtained by also using an amplitude
dependent gain (Pecharroman and Pagola [165]); Appendix 2.
D. First order lag plus integral plus time delay (FOLIPD) model ( G m (s) =
Km e s
)
s(1 + sTm )
m
* Method 1: Method of moments (Astrom and Hagglund [3]).

Method 3: Parameters estimated from the open loop step response and its first and second time derivatives
(Tsang and Rad [109]); Appendix 2.
Method 4: K u and Tu are estimated from estimates of the ultimate and crossover frequencies (Pecharroman
and Pagola [165]) as in Method 6, IPD model.
E.
Second order system plus time delay (SOSPD) model ( G m (s) =
Km e s
)
s + 2 m Tm1s + 1 (1 + Tm1s)(1 + Tm2 s)
K m e s
Tm1
2 2
Method 1: Km , Tm1 , Tm2 , m or Km , Tm1 , m , m assumed known.

Method 2: Parameters estimated using a two-stage identification procedure involving (a) placing a relay in
series with the process in closed loop and (b) placing a proportional controller in series with the
process in closed loop (Sung et al. [139]); Appendix 2.
* Method 3: Parameters obtained in the frequency domain from the ultimate gain, phase and frequency
determined using a relay in series with the closed loop system in a master feedback loop. The
model gain is obtained by the ratio of the integrals (over one period) of the process output to the
controller output. The other parameters are obtained from the frequency domain data (Hwang
[160]).
Method 4: Tm and m estimated from Ku , Tu determined using a relay autotuning method; Km estimated
from the process step response (Hang et al. [35]); Appendix 2.
* Method 5: Parameter estimates back-calculated from discrete time identification method (Ferretti et al.
[161]).
Method 6: Parameteres estimated from the underdamped or overdamped transient response in open loop to a
step input (Jahanmiri and Fallahi [149]); Appendix 2.
* Method 7: Parameters estimated from a least squares time domain method (Lopez et al. [84]).
Method 8: Parameters estimated from data obtained when the process phase lag is 900 and 180 0 ,
respectively (Wang et al. [143]); Appendix 2.
* Method 9: Parameter estimates back-calculated from discrete time identification method (Wang and
Clements [147]).
Method 10: Km , Tm1 and m are determined from the open loop time domain Ziegler-Nichols response
(Shinskey [16], page 151); Tm 2 assumed known.
Method 11: Parameters estimated from two points determined on process frequency response using a relay
and a relay in series with a delay (Tan et al. [39]); Appendix 2.
* Method 12: Parameter estimated back-calculated from discrete time identification method (Lopez et al. [84]).
* Method 13: Parameters estimated from a step response autotuning experiment Honeywell UDC 6000
controller (Astrom et al. [30]).
Method 14: Tm1 = T m2 . m and T m1 obtained from an open loop step test as follows:
Tm1 = 0. 794 ( t 70% t 33% ) , m = 1.937 t 33% 0.937 t 70% . K m assumed known (Miluse et al.
[27b]).
Method 15: K u and Tu are estimated from estimates of the ultimate and crossover frequencies (Pecharroman
and Pagola [165]) as in Method 6, IPD model.
F. Integral squared plus time delay ( I 2PD ) model ( G m (s) =
K me s m
)
s2

G. Second order system (repeated pole) plus integral plus time delay (SOSIPD) model ( G m (s) =
Km e s
s (1 + sTm )
Method 1: K u and Tu are estimated from estimates of the ultimate and crossover frequencies (Pecharroman and
Pagola [165]) as in Method 6, IPD model.
H. Third order system plus time delay (TOLPD) model ( G m (s) =
Km e s
).
(1 + sTm1 )(1 + sTm2 )(1 + sTm 3 )
m
Method 1: Km , Tm1 , Tm2 , Tm3 , m known.
I. Unstable first order lag plus time delay model ( G m (s) =
K me s
1 sTm
Method 1: Km , Tm , m known.
Method 2: The model parameters are obtained by least squares fitting from the open loop frequency
response of the unstable process; this is done by determining the closed loop magnitude and
phase values of the (stable) closed loop system and using the Nichols chart to determine the
open loop response (Huang and Lin [154], Deshpande [164]).
J. Unstable second order system plus time delay model ( G m (s) =
K m e s
)
(1 sTm1 )(1 + sTm 2 )
m
Method 1: Km , Tm1 , Tm2 , m known.

Method 2: The model parameters are obtained by least squares fitting from the open loop frequency
response of the unstable process; this is done by determining the closed loop magnitude and
phase values of the (stable) closed loop system and using the Nichols chart to determine the
open loop response (Huang and Lin [154], Deshpande [164]).
K. Second order system plus time delay model with a positive zero ( G m (s) =
K m (1 sTm3 ) e s
(1 + sTm1 )(1 + sTm2 )
L. Second order system plus time delay model with a negative zero ( G m (s) =
K m (1 + sTm3 )e s
(1 + sTm1 )(1 + sTm 2 )

M. Fifth order system plus delay model ( G m ( s) =
K m (1 + b1s + b2 s2 + b3s3 + b4 s4 + b ss5 ) e s
(1 + a s + a s
Method 1: Km , b1 , b 2 , b 3 , b 4 , b5 , a1 , a2 , a3 , a4 , a5 , m known.
N. General model with a repeated pole ( G m ( s) =
+ a3s 3 + a 4 s4 + a 5s5
K m e s
)
(1 + sTm ) n
m
* Method 1: Strejcs method

O. General stable non-oscillating model with a time delay
P.
Delay model ( G m ( s) = e s m )
Note: * means that the procedure has not been fully described to date.
4. Organisation of the report
The tuning rules are organised in tabular form, as is indicated in the list of tables below. Within each table, the
tuning rules are classified further; the main subdivisions made are as follows:
(i) Tuning rules based on a measured step response (also called process reaction curve methods).
(ii) Tuning rules based on minimising an appropriate performance criterion, either for optimum regulator or
optimum servo action.
(iii) Tuning rules that gives a specified closed loop response (direct synthesis tuning rules). Such rules may be
defined by specifying the desired poles of the closed loop response, for instance, though more generally,
the desired closed loop transfer function may be specified. The definition may be expanded to cover
techniques that allow the achievement of a specified gain margin and/or phase margin.
(iv) Robust tuning rules, with an explicit robust stability and robust performance criterion built in to the design
process.
(v) Tuning rules based on recording appropriate parameters at the ultimate frequency (also called ultimate
cycling methods).
(vi) Other tuning rules, such as tuning rules that depend on the proportional gain required to achieve a quarter
decay ratio or magnitude and frequency information at a particular phase lag.
Some tuning rules could be considered to belong to more than one subdivision, so the subdivisions cannot be
considered to be mutually exclusive; nevertheless, they provide a convenient way to classify the rules. Tuning
rules for the variations that have been proposed in the ideal PI and PID controller structure are included in the
appropriate table. In all cases, one column in the tables summarise the conditions under which the tuning rules
are designed to operate, if appropriate ( Y ( s) = closed loop system output, R( s) = closed loop system input).
Tables 1-3: PI tuning rules FOLPD model - G m ( s) =
K me s
1 + sTm
1
Table 1: Ideal controller G c ( s) = Kc 1 +
. Eighty-five such tuning rules are defined; the references are
Ts
i
(a) Process reaction methods: Ziegler and Nichols [8], Hazebroek and Van der Waerden [9], Astrom
and Hagglund [3], Chien et al. [10], Cohen and Coon [11], Wolfe [12], Murrill [13] page 356,
McMillan [14] page 25, St. Clair [15] page 22 and Shinskey [15a]. Twelve tuning rules are
defined.
(b) Performance index minimisation (regulator tuning): Minimum IAE - Murrill [13] pages 358-363,
Shinskey [16] page 123, ** Shinskey [17], Huang et al. [18], Yu [19]. Minimum ISE - Hazebroek
and Van der Waerden [9], Murrill [13] pages 358-363, Zhuang and Atherton [20], Yu [19].
Minimum ITAE - Murrill [13] pages 358-363, Yu [19]. Minimum ISTSE - Zhuang and Atherton
[20]. Minimum ISTES Zhuang and Atherton [20]. Thirteen tuning rules are defined.
(c) Performance index minimisation (servo tuning): Minimum IAE Rovira et al. [21], Huang et al. [18].
Minimum ISE - Zhuang and Atherton [20], han and Lehman [22]. Minimum ITAE - Rovira et al. [21].
Minimum ISTSE - Zhuang and Atherton [20]. Minimum ISTES Zhuang and Atherton [20]. Seven
tuning rules are defined.
(d) Direct synthesis: Haalman [23], Chen and Yang [23a], Pemberton [24], Smith and Corripio [25], Smith
et al. [26], Hang et al. [27], Miluse et al. [27a], Gorecki et al. [28], Chiu et al. [29], Astrom et al. [30],
Davydov et al. [31], Schneider [32], McAnany [33], Leva et al. [34], Khan and Lehman [22], Hang et
al. [35, 36], Ho et al. [37], Ho et al [104], Tan et al. [39], Voda and Landau [40], Friman and Waller
[41], Smith [42], Cox et al. [43], Cluett and Wang [44], Abbas [45], Bi et al. [46], Wang and Shao
[47]. Thirty-one tuning rules are defined.
(e) Robust: Brambilla et al. [48], Rivera et al. [49], Chien [50], Thomasson [51], Fruehauf et al. [52],
Chen et al. [53], Ogawa [54], Lee et al. [55], Isaksson and Graebe [56], Chun et al. [57]. Ten tuning
rules are defined.
(f) Ultimate cycle: McMillan [58], Shinskey [59] page 167, **Shinskey [17], Shinskey [16] page
148, Hwang [60], Hang et al. [65], Zhuang and Atherton [20], Hwang and Fang [61]. Twelve tuning
rules are defined.
1
Table 2: Controller G c ( s) = Kc b +
Two direct synthesis tuning rules are defined by Astrom and Hagglund
T
i s
[3] - page 205-208.
1
E (s ) Kc R( s) . One performance index
Table 3: Two degree of freedom controller: U(s) = K c 1 +
Tis
minimisation tuning rule is defined by Taguchi and Araki [61a].

Tables 4-7: PI tuning rules - non-model specific
1
Table 4: Controller G c ( s) = Kc 1 +
. Nineteen such tuning rules are defined; the references are:
Ts
i
(a) Ultimate cycle: Ziegler and Nichols [8], Hwang and Chang [62], ** Hang et al. [36], McMillan [14]
page 90, Pessen [63], Astrom and Hagglund [3] page 142, Parr [64] page 191, Yu [122] page
11. Seven tuning rules are defined.
(b) Other tuning rules: Parr [64] page 191, McMillan [14] pages 42-43, Parr [64] page 192,
Hagglund and Astrom [66], Leva [67], Astrom [68], Calcev and Gorez [69], Cox et al. [70]. Eight
(c) Direct synthesis: Vrancic et al. [71], Vrancic [72], Friman and Waller [41], Kristiansson and
Lennartson [158a]. Four tuning rules are defined.

1
Table 5: Controller G c ( s) = K c b +
. One direct synthesis tuning rule is defined by Astrom and Hagglund
Ti s
[3] page 215.
1
Table 6: Controller U( s) = Kc 1 +
E( s) + Kc ( b 1)R (s) . One direct synthesis tuning rule is defined by
T
is
Vrancic [72].
Table 7: Controller U( s) = Kc Y(s)
Kc
E ( s) . One direct synthesis tuning rule is defined by Chien et al. [74].
Ti s
K me s m
s
Tables 8-11: PI tuning rules IPD model G m ( s) =
1
. Twenty such tuning rules are defined; the references are:
Ts
i
(a) Process reaction methods: Ziegler and Nichols [8], Wolfe [12], Tyreus and Luyben [75], Astrom and
Hagglund [3] page 138. Four tuning rules are defined.
(b) Regulator tuning performance index minimisation: Minimum ISE Hazebroek and Van der
Waerden [9]. Minimum IAE - Shinskey [59] page 74. Minimum ITAE - Poulin and Pomerleau [82].
Four tuning rules are defined.
(c) Ultimate cycle: Tyreus and Luyben [75], ** Shinskey [17]. Two tuning rules are defined.
(d) Robust: Fruehauf et al. [52], Chien [50], Ogawa [54]. Three tuning rules are defined.
(e) Direct synthesis: Wang and Cluett [76], Cluett and Wang [44], Rotach [77], Poulin and Pomerleau
[78], Kookos et al. [38]. Five tuning rules are defined.
(f) Other methods: Penner [79], Srividya and Chidambaram [80]. Two tuning rules are defined.
1 1
. One robust tuning rule is defined by Tan et al. [81].
Ts
i 1 + Tf s
Table 10: Controller U( s) = Kc Y(s)
Kc
E ( s) . One direct synthesis tuning rule is defined by Chien et al. [74].
Ti s
1
E (s ) Kc R( s) . Two performance index
Tis
minimisation - servo/regulator tuning rules are defined by Taguchi and Araki [61a] and Pecharroman
and Pagola [134b].
Tables 12-14: PI tuning rules FOLIPD model G m (s) =
Km e s
s(1 + sTm )
m
1
. Six such tuning rules are defined; the references are:
Ts
i
(a) Ultimate cycle: McMillan [58]. One tuning rule is defined.
(b) Regulator tuning minimum performance index: Minimum IAE Shinskey [59] page 75. Shinskey
[59] page 158. Minimum ITAE - Poulin and Pomerleau [82]. Four tuning rules are defined.
(c) Direct synthesis - Poulin and Pomerleau [78]. One tuning rule is defined.
1
Table 13: Controller G c ( s) = Kc b +
. One direct synthesis tuning rule is defined by Astrom and Hagglund
Ti s
[3] pages 210-212.
1
Tis
minimisation tuning rules are defined by Taguchi and Araki [61a] and Pecharroman and Pagola [134b].
Tables 15-16: PI tuning rules SOSPD model
K m e s
Tm1 s2 + 2 m Tm1s + 1
2
Km e s
(1 + Tm1s)(1 + Tm2 s)
m
or

1
. Ten tuning rules are defined; the references are:
Ts
i
(a) Robust: Brambilla et al. [48]. One tuning rule is defined.
(b) Direct synthesis: Tan et al. [39]. One tuning rule is defined.
(c) Regulator tuning minimum performance index: Minimum IAE - Shinskey [59] page 158, **
Shinskey [17], Huang et al. [18], Minimum ISE McAvoy and Johnson [83], Minimum ITAE
Lopez et al. [84]. Five tuning rules are defined.
(d) Servo tuning minimum performance index: Minimum IAE - Huang et al. [18]. One tuning rule is
defined.
(e) Ultimate cycle: Hwang [60], ** Shinskey [17]. Two tuning rules are defined.
1
E (s ) Kc R( s) . Three performance index
T
is
minimisation tuning rules are defined by Taguchi and Araki [61a] and Pecharroman and Pagola [134a],
[134b].
Table 17: PI tuning rules SOSIPD model (repeated pole) G m (s) =
Km e s
s (1 + sTm )
1
Tis
minimisation tuning rules are defined by Taguchi and Araki [61a] and Pecharroman and Pagola [134b].
Km e s
(1 + sTm1 )(1 + sTm2 )(1 + sTm 3 )
m
Tables 18-19: PI tuning rules third order lag plus delay (TOLPD) model
1
. One *** tuning rule is defined. The reference is Hougen [85].
Ts
1
E (s ) Kc R( s) . One performance index
Tis
minimisation tuning rule is defined by Taguchi and Araki [61a].

K me s
1 sTm
Table 20: PI tuning rules - unstable FOLPD model
1
. Six tuning rules are defined; the references are:
Ts
i
(a) Direct synthesis: De Paor and OMalley [86], Venkatashankar and Chidambaram [87], Chidambaram
[88], Ho and Xu [90]. Four tuning rules are defined.
(b) Robust: Rotstein and Lewin [89]. One tuning rule is defined.
(c) Ultimate cycle: Luyben [91]. One tuning rule is defined.
K m e s
(1 sTm1 )(1 + sTm 2 )
m
Table 21: PI tuning rules - unstable SOSPD model
1
. Three tuning rules are defined; the references are:
Ts
i
(a) Ultimate cycle: McMillan [58]. One tuning rule is defined.
(b) Minimum performance index regulator tuning: Minimum ITAE Poulin and Pomerleau [82]. Two
Table 22: PI tuning rules delay model e s m
1
. Two tuning rules are defined; the references are:
Ts
(a) Direct synthesis: Hansen [91a].

(b) Minimum performance index regulator tuning: Shinskey [57].
K m e s
1 + sTm
Tables 23-40: PID tuning rules - FOLPD model
1
+ Td s . Fifty-seven tuning rules are defined; the references are:
Ti s
(a) Process reaction: Ziegler and Nichols [8], Astrom and Hagglund [3] page 139, Parr [64] page
194, Chien et al. [10], Murrill [13]- page 356, Cohen and Coon [11], Astrom and Hagglund [93]pages 120-126, Sain and Ozgen [94]. Eight tuning rules are defined.
(b) Minimum performance index regulator tuning: Minimum IAE Murrill [13] pages 358-363,
Cheng and Hung [95]. Minimum ISE - Murrill [13] pages 358-363, Zhuang and Atherton [20].
Minimum ITAE - Murrill [13] pages 358-363. Minimum ISTSE - Zhuang and Atherton [20].
Minimum ISTES - Zhuang and Atherton [20]. Minimum error - step load change - Gerry [96]. Eight
(c) Minimum performance index servo tuning: Minimum IAE - Rovira et al. [21], Wang et al. [97].
Minimum ISE - Wang et al. [97], Zhuang and Atherton [20]. Minimum ITAE - Rovira et al. [21],
Cheng and Hung [95], Wang et al. [97]. Minimum ISTSE Zhuang and Atherton [20]. Minimum
ISTES Zhuang and Atherton [20]. Nine tuning rules are defined.
(d) Ultimate cycle: Pessen [63], Zhuang and Atherton [20], Pi-Mira et al. [97a], Hwang [60], Hwang and
Fang [61], McMillan [58], Astrom and Hagglund [98], Li et al. [99], Tan et al. [39], Friman and
Waller [41]. Fourteen tuning rules are defined.
(e) Direct synthesis: Gorecki et al. [28], Smith and Corripio [25], Suyama [100], Juang and Wang [101],
Cluett and Wang [44], Zhuang and Atherton [20], Abbas [45], Camacho et al.[102, Ho et al. [103],
Ho et al [104], Morilla et al. [104a]. Fourteen tuning rules are defined.
(f) Robust: Brambilla et al. [48], Rivera et al. [49], Fruehauf et al. [52], Lee et al. [55]. Four tuning rules.
1
1
Table 24: Ideal controller with first order filter G c ( s) = Kc 1 +
+ Td s
. Three robust tuning rules are
Ti s
Tf s + 1
defined by ** Morari and Zafiriou [105], Horn et al. [106] and Tan et al. [81].
1
1 + b1s
Table 25: Ideal controller with second order filter G c ( s) = Kc 1 +
+ Td s
. One robust tuning
2
Ti s
1 + a 1s + a 2s
rule is defined by Horn et al. [106].
1
Table 26: Ideal controller with set-point weighting G c ( s) = Kc b +
+ Td s . One direct synthesis tuning rule
Ti s
is defined by Astrom and Hagglund [3] pages 208-210.

Table 27. Ideal controller with first order filter and set-point weighting:
1
1 + 0.4Trs
U(s) = K c 1 +
+ Td s
Y( s) . One direct synthesis tuning rule is defined
R (s )
1 + sTr
Tis
Tf s + 1
by Normey-Rico et al. [106a].
1 1 + sTd
Table 28: Classical controller G c ( s) = K c 1 +

. Twenty tuning rules are defined; the references are:
T
Ti s
1+ s d
N
(a) Process reaction: Hang et al. [36] page 76, Witt and Waggoner [107], St. Clair [15] page 21,
Shinskey [15a]. Five tuning rules are defined.
(b) Minimum performance index regulator tuning: Minimum IAE - Kaya and Scheib [108], Witt and
Waggoner [107]. Minimum ISE - Kaya and Scheib [108]. Minimum ITAE - Kaya and Scheib [108],
Witt and Waggoner [107]. Five tuning rules are defined.
(c) Minimum performance index servo tuning: Minimum IAE - Kaya and Scheib [108], Witt and
Waggoner [107]. Minimum ISE - Kaya and Scheib [108]. Minimum ITAE - Kaya and Scheib [108],
Witt and Waggoner [107]. Five tuning rules are defined.
(d) Direct synthesis: Tsang and Rad [109], Tsang et al. [111]. Two tuning rules are defined.
(e) Robust: Chien [50]. One tuning rule is defined.
(f) Ultimate cycle: Shinskey [59] page 167, Shinskey [16] page 143. Two tuning rules are defined.
1
Td s
. Two tuning rules are defined; the
E( s)
Table 29: Non-interacting controller U (s) = K c 1 +
Y
(
s
)
Td s
Tis
1+
references are:
(a) Minimum performance index regulator tuning: Minimum IAE - Huang et al. [18].
(b) Minimum performance index servo tuning: Minimum IAE - Huang et al. [18].
1
Td s
Table 30: Non-interacting controller U( s) = Kc 1 +
Y(s) . Five tuning rules are defined; the
E ( s)
sT
Ti s
1+ d
N
references are:
(a) Minimum performance index servo tuning: Minimum ISE - Zhuang and Atherton [20], Minimum
ISTSE - Zhuang and Atherton [20]. Minimum ISTES - Zhuang and Atherton [20]. Three tuning
rules are defined.
(b) Ultimate cycle: Zhuang and Atherton [20], Shinskey [16] page 148. Two tuning rules are defined.
Td s
1
E (s)
Table 31: Non-interacting controller U (s) = K c +
Y ( s) . Six tuning rules are defined; the
sTd
T
s
i
1+
N
references are:
(a) Minimum performance index regulator tuning: Minimum IAE Kaya and Scheib [108]. Minimum
ISE Kaya and Scheib [108]. Minimum ITAE Kaya and Scheib [108].
(b) Minimum performance index servo tuning: Minimum IAE Kaya and Scheib [108]. Minimum ISE
Kaya and Scheib [108]. Minimum ITAE Kaya and Scheib [108].
Table 32: Non-interacting controller with setpoint weighting:
1
Kc Td s
U( s) = Kc b +
Y( s) + Kc ( b 1)Y( s) . Three ultimate cycle tuning rules are
E (s)
Ti s
1 + Td s N
defined by Hang and Astrom [111], Hang et al. [65] and Hang and Cao [112].
1
1 + Td s
Table 33: Industrial controller U( s) = Kc 1 +
Y(s) . Six tuning rules are defined: the
R( s)
Ts
Ti s
1+ d
N
reference are:
(a) Minimum performance index regulator tuning: Minimum IAE - Kaya and Scheib [108]. Minimum
ISE - Kaya and Scheib [108]. Minimum ITAE - Kaya and Scheib [108]. Three tuning rules are
defined.
(b) Minimum performance index servo tuning: Minimum IAE - Kaya and Scheib [108]. Minimum ISE Kaya and Scheib [108]. Minimum ITAE - Kaya and Scheib [108]. Three tuning rules are defined.
1
Table 34: Series controller Gc ( s) = Kc 1 +
(1 + sTd ) . Three tuning rules are defined; the references are:
Ti s
(a) ******: Astrom and Hagglund [3] page 246.

(b) Ultimate cycle: Pessen [63].
(c) Direct synthesis: Tsang et al. [110].
1
sTd
. One robust tuning rule is
Table 35: Series controller with filtered derivative Gc ( s) = Kc 1 +
1 +
sT
Ti s
1 + d
N
defined by Chien [50].
1
Td s
. Three tuning rules are defined; the
Table 36: Controller with filtered derivative G c ( s) = Kc 1 +
+
Ti s 1 + s Td
N
references are:
(a) Robust: Chien [50], Gong et al. [113]. Two tuning rules are defined.
(b) Direct synthesis: Davydov et al. [31]. One tuning rule is defined.
1
Table 37: Alternative non-interacting controller 1 - U( s) = Kc 1 +
E( s) Kc Td sY ( s) . Six ultimate cycle
Ti s
tuning rules are defined; the references are: Shinskey [59] page 167, ** Shinskey [17], Shinskey [16]
page 143, VanDoren [114].
2
1 1 + 05
. m s + 0.0833 m s 2
Table 38: Alternative filtered derivative controller - G c ( s) = Kc 1 +
. One direct
Ti s
[1 + 01. ms]
synthesis tuning rule is defined by Tsang et al. [110].

Table 39: I-PD controller U(s) =
Kc
E (s) K c (1 + Td s)Y(s) . Two direct synthesis tuning rules are defined by Chien
Ti s
et al. [74] and Argelaguet et al. [114a].
1
T
s
T
s
d
d
E( s) K c +
R (s) . One
Table 40: Two degree of freedom controller: U (s) = K c 1 +
+
Tis
Td
Td
1+
s
1+
s
N
N
performance index minimisation tuning rule is defined by Taguchi and Araki [61a].
Tables 41-48: PID tuning rules - non-model specific
1
+ Td s . Twenty five tuning rules are defined; the references are
Ti s
(a) Ultimate cycle: Ziegler and Nichols [8], Blickley [115], Parr [64] pages 190-191, De Paor [116],
Corripio [117] page 27, Mantz andTacconi [118], Astrom and Hagglund [3] page 142, Astrom
and Hagglund [93], Atkinson and Davey [119], ** Perry and Chilton [120], Yu [122] page 11, Luo
et al. [121], McMillan [14] page 90, McAvoy and Johnson [83], Karaboga and Kalinli [123], Hang
and Astrom [124], Astrom et al. [30], St. Clair [15] - page 17, Shin et al. [125]. Nineteen tuning rules
are defined.
(b) Other tuning: Harriott [126], Parr [64] pages 191, 193, McMillan [14] - page 43, Calcev and Gorez
[69], Zhang et al. [127], Garcia and Castelo [127a]. Six tuning rules are defined.
1
Td s
. Eight tuning rules are defined; the
Table 42: Controller with filtered derivative Gc ( s) = Kc 1 +
+
Ti s 1 + Td s
N
references are:
(a) Direct synthesis: Vrancic [72], Vrancic [73], Lennartson and Kristiansson [157], Kristiansson and
Lennartson [158], Kristiansson and Lennartson [158a]. Six tuning rules are defined.
(b) Other tuning: Leva [67], Astrom [68]. Two tuning rules are defined.
Table 43: Ideal controller with set-point weighting:
1
U( s) = Kc Fp R(s) Y(s) +
( Fi R(s) Y(s)) + Td s( Fd R(s) Y(s)) . One ultimate cycle tuning rule is
Ts
i
defined by Mantz and Tacconi [118].
1
Table 42: Ideal controller with proportional weighting G c ( s) = Kc b +
+ Td s . One direct synthesis tuning
Ti s
rule is defined by Astrom and Hagglund [3] page 217.
1
K Ts
E( s) c d Y( s) . One ultimate cycle tuning rule is
sT
Ti s
1+ d
N
defined by Fu et al. [128].

1
Table 45: Series controller U( s) = Kc 1 +
(1 + sTd ) . Three ultimate cycle tuning rules are defined by Pessen
Ti s
[131], Pessen [129] and Grabbe et al. [130].
1
sTd
. One ultimate cycle tuning
Table 46: Series controller with filtered derivative U( s) = Kc 1 +
1 +
sT
Ti s
1 + d
N
rule is defined by Hang et al. [36] - page 58.
1 1 + sTd
Table 47: Classical controller U( s) = Kc 1 +
. One ultimate cycle tuning rule is defined by Corripio
T
Ti s
1+ s d
N
[117].
1
E ( s) Kc Td sY (s) . One ultimate cycle tuning rule is
Ti s
defined by VanDoren [114].

K m e s m
s
1
Table 49: Ideal controller Gc ( s) = K c 1 +
+ Td s . Five tuning rules are defined; the references are:
T
s
i
Tables 49-58: PID tuning rules - IPD model
(a) Process reaction: Ford [132], Astrom and Hagglund [3] page 139. Two tuning rules are defined.
(b) Direct synthesis: Wang and Cluett [76], Cluett and Wang [44], Rotach [77]. Three tuning rules are
defined.
Table 50: Ideal controller with first order filter, set-point weighting and output feedback:
1
1 + 0.4Trs
U(s ) = K c 1 +
+ Tds
Y(s ) K 0 Y(s) . One direct synthesis tuning rule
R( s)
Ti s
1 + sTr
Tf s + 1
has been defined by Normey-Rico et al. [106a].
1
Td s
. One robust tuning rule has
Table 51: Ideal controller with filtered derivative G c ( s) = Kc 1 +
+
sTd
Ts
i
1+
N
been defined by Chien [50].
1
Td s
Table 52: Series controller with filtered derivative G c ( s) = Kc 1 +
1 +
Ts
Ts
i
1 + d
N
1 1 + Td s
. Five tuning rules have been defined; the references
Table 53: Classical controller G c ( s) = Kc 1 +
Ti s 1 + Td s
N
are:
(a) Ultimate cycle: Luyben [133], Belanger and Luyben [134]. Two tuning rules have been defined.
(b) Robust: Chien [50]. One tuning rule has been defined.
(c) Performance index minimisation regulator tuning: ** Minimum IAE - Shinskey [17], Shinskey [59]
page 74. Two tuning rules have been defined.
1
Table 54: Alternative non-interacting controller 1: U( s) = Kc 1 +
E( s) Kc Td sY ( s) . Two performance index
T
is
minimisation rules minimum IAE regulator tuning have been defined by Shinskey [59] page 74 and
** Shinskey [17].
Table 55: I-PD controller U(s) =
Kc
E (s) K c (1 + Td s)Y(s) . One direct synthesis tuning rule has been defined by
Ti s
Chien et al. [74].
1
) E( s) + Kc ( b 1) R (s) Kc TdsY( s) . One direct synthesis tuning rule has
Tis
been defined by Hansen [91a].
1
T
s
T
s
d
d
E( s) K c +
R (s) . Two
+
Tis
Td
Td
1
+
s
1
+
s
N
N
minimum performance index servo/regulator tuning have been defined by Taguchi and Araki [61a] and
Pecharroman and Pagola [134a].
Table 56: Controller U(s) = Kc (1 +
Km e s
s(1 + sTm )
m
Tables 58-67: PID tuning rules - FOLIPD model
1
+ Td s . One ultimate cycle tuning rule has been defined by Millan
Ti s
[58].
1
1
Table 59: Ideal controller with filter G c ( s) = Kc 1 +
+ Td s
. Three robust tuning rules have been
T
s
1 + Tf s
i
defined by Tan et al. [81], Zhang et al. [135] and Tan et al. [136].
1
Table 60: Ideal controller with set-point weighting G c ( s) = Kc b +
+ Td s . One direct synthesis tuning rule
Ti s
has been defined by Astrom and Hagglund [3] - pages 212-213.
1 1 + Td s
. Five tuning rules have been defined; the references
Ti s 1 + Td s
N
are as follows:
(a) Robust: Chien [50]. One tuning rule is defined.
(b) Minimum performance index regulator tuning: Minimum IAE Shinskey [59] page 75, Shinskey
[59] pages 158-159, Minimum ITAE - Poulin and Pomerleau [82], [92]. Four tuning rules are
defined.
1
Td s
Table 62: Series controller with derivative filtering G c ( s) = Kc 1 +
1 +
Ts
Ts
i
1 + d
N
1
U(s) = Kc 1 + E (s) K c Td sY ( s) . Two minimum
Ti s
performance index (minimum IAE) regulator tuning rules have been defined by Shinskey [59] page
75, page 159.
1
Td s
Table 64: Ideal controller with filtered derivative: G c ( s) = Kc 1 +
+
Ti s 1 + s Td
N
Table 63: Alternative non-interacting controller 1:
Kc
[ Fi R(s) Y(s)] + Kc Td s[ Fd R (s) Y(s)] . One ultimate cycle tuning rule
Ti s
has been defined by Oubrahim and Leonard [138].
1+ T s 1 + T s
i
d
. One direct synthesis tuning rule has

Table 66: Alternative classical controller: G c ( s) = Kc
1 + Td s 1 + Td s
N
N
been defined by Tsang and Rad [109].
U(s) = Kc Fp R ( s) Y( s) +
Table 67: Two degree of freedom controller:
1
T
s
T
s
d
d
E( s) K c +
R (s) . Two minimum performance index
U (s) = K c 1 +
+
Tis
Td
Td
1+
s
1+
s
N
N
servo/regulator tuning have been defined by Taguchi and Araki [61a] and Pecharroman and Pagola
[134a].
Km e s
K m e s
or
2 2
(1 + sTm1 )(1 + sTm 2 ) Tm1 s + 2 m Tm1s + 1
m
Tables 68-79: PID tuning rules - SOSPD model
1
+ Td s . Twenty seven tuning rules have been defined; the
Ti s
references are:
(a) Minimum performance index servo tuning: Minimum ITAE Sung et al. [139]. One tuning rule is
defined.
(b) Minimum performance index regulator tuning: Minimum ITAE Sung et al. [139], Lopez et al.
[84]. One tuning rule is defined.
(c) Ultimate cycle: Hwang [60], Shinskey [16] page 151. Three tuning rules are defined.
(d) Direct synthesis: Hang et al. [35], Ho et al. [140], Ho et al. [141], Ho et al. [142], Wang et al. [143],
Leva et al. [34], Wang and Shao [144], Pemberton [145], Pemberton [24], Suyama [100], Smith et al.
[146], Chiu et al. [29], Wang and Clemens [147], Gorez and Klan [147a], Miluse et al. [27a], Miluse et
al. [27b], Seki et al. [147b], Landau and Voda [148]. Nineteen tuning rules are defined.
(e) Robust: Brambilla et al. [48], Chen et al. [53], Lee et al. [55]. Three tuning rules are defined.
1
1
Table 69: Filtered controller G c ( s) = Kc 1 +
+ Td s
. One robust tuning rule has been defined by Hang
Ti s
Tf s + 1
et al. [35].
b s+1
1
Table 70: Filtered controller G c ( s) = Kc 1 +
+ Td s 1
. One robust tuning rule has been defined by
T
s
a 1s + 1
i
Jahanmiri and Fallahi [149].
1 1 + Td s
. Seven tuning rules have been defined; the
Ti s 1 + Td s
N
references are:
(a) Minimum performance index regulator tuning: Minimum IAE - Shinskey [59] page 159, **
Shinskey [59], ** Shinskey [17], ** Shinskey [17]. Minimum ISE McAvoy and Johnson [83]. Five
(b) Direct synthesis: Astrom et al. [30], Smith et al. [26]. Two tuning rules are defined.
1 1 + NTd s
Table 72: Alternative classical controller G c ( s) = Kc 1 +
. One ***** tuning rule has been

Ti s 1 + Td s
defined by Hougen [85].

1
Table 73: Alternative non-interacting controller 1: U( s) = K c 1 +
E ( s) Kc Td sY ( s) . Three minimum
Ti s
performance index (minimum IAE) regulator tuning rules have been defined by Shinskey [59] page
158, ** Shinskey [17], ** Shinskey [17].
1
Table 74: Series controller G c ( s) = K c 1 +
(1 + Td s) . One minimum performance index - regulator tuning rule
Ti s
has been defined by Haalman [23].
1
Td s
. Two tuning rules have been
E( s)
Table 75: Non-interacting controller U (s) = K c 1 +
Y
(
s
)
Td s
Tis
1+
defined. The references are:

(a) Minimum performance index regulator tuning: Minimum IAE - Huang et al. [18].
(b) Minimum performance index servo tuning: Minimum IAE - Huang et al. [18].
K
U(s) = Kc Fp R ( s) Y( s) + c [ Fi R( s) Y(s) ] + Kc Td s[ Fd R ( s) Y( s)] . One ultimate cycle tuning rule
Ti s
has been defined by Oubrahim and Leonard [138].
1
Table 77: Non-interacting controller U( s) = Kc b + [ R (s) Y(s) ] ( c + Tds)Y( s) . One direct synthesis tuning
Ts
i
rule has been defined by Hansen [150].
1
Td s
. Two tuning rules are defined;
+
sT
Ts
i
1 + d
N
the references are:
(a) Direct synthesis: Hang et al. [151].
(b) Robust: Hang et al. [151].
1
T
s
T
s
d
d
E( s) K c +
R (s) . Three
+
Tis
Td
Td
1
+
s
1
+
s
N
N
minimum performance index servo/regulator tuning rules have been defined by Taguchi and Araki
[61a] and Pecharroman and Pagola [134a], [134b].
Table 80: PID tuning rules - I 2PD model G m (s) =
K me s m
s2
1
) E( s) + Kc ( b 1) R (s) Kc TdsY( s) . One direct synthesis tuning rule has
Tis
been defined by Hansen [91a].
Table 80: Controller U(s) = Kc (1 +
Table 81: PID tuning rules SOSIPD model (repeated pole) G m (s) =
Km e s
s (1 + sTm )
1
T
s
T
s
d
d
E( s) K c +
R (s) . Two
+
Tis
Td
Td
1+
s
1+
s
N
N
minimum performance index servo/regulator tuning rules have been defined by Taguchi and Araki
[61a] and Pecharroman and Pagola [134a].
Tables 82-84: PID tuning rules - SOSPD model with a positive zero
K m (1 sTm3 ) e s
(1 + sTm1 )(1 + sTm2 )
1
Td s
. One robust tuning rule has been
+
Ti s 1 + Td s
N
1 1 + Td s
. One robust tuning rule has been defined by Chien
Ti s 1 + Td s
N
[50].
1
Td s
1 +
Ts
Ts
i
1 + d
N
Tables 85-88: PID tuning rules - SOSPD model with a negative zero
K m (1 + sTm 3 ) e s
(1 + sTm1 )(1 + sTm 2 )
1
+ Td s . One minimum performance index tuning rule has been
Ti s
defined by Wang et al. [97].
1
Td s
. One robust tuning rule has been
+
Ti s 1 + Td s
N
1 1 + Td s
. One robust tuning rule has been defined by Chien
Ti s 1 + Td s
N
[50].
1
Td s
1 +
Ts
Ts
i
1 + d
N
Km e s
(1 + sTm1 )(1 + sTm 2 )(1 + sTm 3 )
m
Table 89-90: PID tuning rules - TOLPD model
1
+ Td s . Two minimum performance index tuning rules have been
Ti s
defined by Polonyi [153].
1
T
s
T
s
d
d
E( s) K c +
R (s) .One
+
Tis
Td
Td
1+
s
1+
s
N
N
minimum performance index tuning rule has been defined by Taguchi and Araki [61a].
K m e s
1 sTm
Tables 91-93: PID tuning rules - unstable FOLPD model
1
+ Td s . Three direct synthesis tuning rules are defined by De Paor
Ti s
and OMalley [86], Chidambaram [88] and Valentine and Chidambaram [154].
1
KTs
E ( s) c d Y(s) . Two tuning rules have been
sT
Ti s
1+ d
N
defined; the references are:
(a) Minimum performance index servo tuning: Minimum IAE - Huang and Lin [155]
(b) Minimum performance index servo tuning: Minimum IAE - Huang and Lin [155]
1 1 + Td s
. One performance index minimisation regulator
Ti s 1 + Td s
N
tuning rule has been defined by Shinskey [16] page 381.
Tables 94-97: PID tuning rules - unstable SOSPD model G m ( s) =
Km e s
(1 sTm1 )(1 + sTm 2 )
m
1
+ Td s . Two tuning rules have been defined; the references are
Ti s
(a) Ultimate cycle: McMillan [58]

(b) Robust: Rotstein and Lewin [89].
1 1 + Tds
. Two minimum performance index tuning rules
Ti s 1 + Td s
N
(regulator - minimum ITAE) have been defined by Poulin and Pomerleau [82], [92].
1
Table 96: Series controller Gc ( s) = Kc 1 +
(1 + Td s) . One direct synthesis tuning rule has been defined by
Ti s
Ho and Xu [90].
Table 97: Non-interacting controller
1
KTs
U( s) = Kc 1 + E ( s) c d Y(s) . Two tuning rules have been
sT
Ti s
1+ d
N
defined; the references are

(a) Minimum performance index servo tuning: Minimum IAE - Huang and Lin [155]
(b) Minimum performance index regulator tuning: Minimum IAE - Huang and Lin [155]
Table 98: PID tuning rules general model with a repeated pole G m ( s) =
K m e s
(1 + sTm ) n
m
1
+ Td s . One direct synthesis tuning rule has been defined by
Ti s
Skoczowski and Tarasiejski [156]
Table 99: PID tuning rules general stable non-oscillating model with a time delay
1
+ Td s . One direct synthesis tuning rule has been defined by Gorez
Ti s
and Klan [147a].
Tables 100-101: PID tuning rules fifth order model with delay
K (1 + b1s + b2 s2 + b3s3 + b4 s4 + b ss5 ) e s m
G m ( s) = m
1 + a1s + a2 s2 + a3s 3 + a 4 s4 + a 5s5
1
+ Td s . One direct synthesis tuning rule is defined by Vrancic et
Ti s
al. [159].
1
Tds
. One direct synthesis tuning rule is
+
Td s
Ts
i
1+
N
defined by Vrancic et al. [159].
** some more information needed.

The number of tuning rules in each table is included in the data. Servo and regulator tuning rules are counted
separately; otherwise, rules in which different tuning parameters are provided for a number of variations in
process parameters or desired response parameters (such as desired gain margin, phase margin or closed loop
response time constant) are counted as one tuning rule. Tabular summaries are provided below.
Table A: Model structure and tuning rules a summary for PI controllers
Model
Stable
FOLPD
Non-model
specific
IPD
FOLIPD
SOSPD
SOSIPD
TOLPD
Unstable
FOLPD
Unstable
SOSPD
Delay
model
TOTAL
Process
reaction
Direct
Synthesis
Ultimate
cycle
Robust
tuning
Other rules
Total
12
Minimise
Performanc
e index
21
33
10
12
88 (53%)
23 (14%)
4
0
0
0
0
0
6
6
9
2
1
0
6
2
1
0
0
4
2
1
2
0
0
1
4
0
1
0
0
1
2
0
0
0
1
0
24 (14%)
9 (5%)
13 (7%)
2 (1%)
2 (1%)
6 (4%)
3 (2%)
2 (1%)
16
48
54
24
18
11
171
Table B: Model structure and tuning rules a summary for PID controllers
Model
Stable
FOLPD
Non-model
specific
IPD
FOLIPD
SOSPD
I 2PD
SOSIPD
SOSPD
pos. zero
SOSPD
neg. zero
TOLPD
Unstable
FOLPD
Unstable
SOSPD
Higher
order
TOTAL
Process
reaction
Direct
Synthesis
Ultimate
cycle
Robust
tuning
Other rules
Total
13
Minimise
Performanc
e index
45
24
28
12
123 (44%)
27
42 (15%)
2
0
0
0
6
8
16
0
6
2
23
1
2
2
4
0
3
6
6
0
0
0
1
0
19 (7%)
18 (6%)
50 (17%)
1 (0%)
0
0
2
0
0
0
0
0
0
3
0
0
2 (1%)
3 (1%)
4 (1%)
0
0
3
3
0
3
0
0
0
0
0
0
3 (1%)
6 (2%)
7 (2%)
4 (1%)
15
88
71
64
34
10
282
Table C: Model structure and tuning rules a summary for PI/PID controllers
Process
reaction
Model
Stable
FOLPD
Non-model
specific
IPD
FOLIPD
SOSPD
I 2PD
SOSIPD
SOSPD
pos. zero
SOSPD
neg. zero
TOLPD
Unstable
FOLPD
Unstable
SOSPD
Delay
model
Higher
order
TOTAL
Direct
Synthesis
Ultimate
cycle
Robust
tuning
Other rules
Total
25
Minimise
Performanc
e index
66
57
38
24
211 (47%)
14
34
16
65 (15%)
6
0
0
0
12
14
25
0
12
4
24
1
4
3
6
0
7
6
7
0
2
0
1
0
43 (9%)
27 (6%)
63 (14%)
1 (0%)
0
0
4
0
0
0
0
0
0
3
0
0
4 (1%)
3 (1%)
4 (1%)
0
0
4
3
0
7
0
0
0
1
1
0
5 (1%)
12 (3%)
10 (2%)
2 (0%)
4 (1%)
31
136
125
88
52
21
453
Table D: PI controller structure and tuning rules a summary

Stable
FOLPD
Nonmodel
specific
IPD
1
G c ( s) = Kc 1 +
Ts
85
19
20
1
G c ( s) = Kc b +
T
i s
1
E (s ) Kc R( s)
U(s) = K c 1 +
Tis
Controller structure
FOLIP
D
SOSPD
Other
Total
10
12
152
(92%)
6
(4%)
10
(4%)
2
(1%)
1 1
G c ( s) = Kc 1 +
Ts
i 1 + Tf s
Total
88
21
23
12
15
1
(0%)
167
U( s) = Kc Y(s)
Kc
E ( s)
Ti s
Table E: PID controller structure and tuning rules a summary

Stable
FOLPD
Nonmodel
specific
IPD
57
25
1
1
G c ( s) = Kc 1 +
+ Td s
Ti s
Tf s + 1
b s+1
1
G c ( s) = Kc 1 +
+ Td s 1
Ti s
a 1s + 1
FOLIP
D
SOSPD
Other
Total
27
11
126
(45%)
7
(3%)
1
(0%)
1
1 + b1s
G c ( s) = Kc 1 +
+ Td s
T
s
1
+
a 1s + a 2s 2
3
(1%)
1
G c ( s) = Kc b +
+ Td s
Ti s
3
(1%)
Subtotal
67
34
31
14
158
(56%)
1 1 + sTd
G c ( s) = K c 1 +
T
Ti s
1+ s d
N
20
43
(15%)
1+ T s 1 + T s
i
d
G c ( s) = Kc
1 + Td s 1 + Td s
N
N
1
(0%)
1 1 + NTd s
G c ( s) = Kc 1 +
i s 1 + Td s
1
(0%)
1
Gc ( s) = Kc 1 +
(1 + sTd )
Ti s
8
(3%)
6
(2%)
1 1 + 05
. m s + 0.0833 m s 2
G c (s) = K c 1 +
2
Ti s
[1 + 01. ms]
1
(0%)
Subtotal
25
60
(22%)
5
(2%)
6
(2%)
1
G c ( s) = Kc 1 +
+ Td s
T
s
1
Td s
G c ( s) = Kc 1 +
+
Ti s 1 + s Td
1
sTd
Gc ( s) = Kc 1 + 1 +
sT
Ti s
1+ d
1
Td s
U( s) = Kc 1 + E ( s)
Y(s)
sTd
T
s
i
1+
N
1
Td s
E( s)
U (s) = K c 1 +
Y
(
s
)
Td s
Tis
1+
18
(6%)
Stable
FOLPD
Nonmodel
specific
IPD
Td s
1
E (s)
U (s) = K c +
Y ( s)
sTd
T
s
i
1+
N
1
K Ts
U( s) = Kc 1 +
E( s) c d Y( s)
sT
T
s
i
1+ d
N
FOLIP
D
SOSPD
Other
Total
6
(2%)
6
(2%)
14
(5%)
1
(0%)
3
(1%)
E(s ) K + Tds R (s )
c
T
s
1+ d s
9
(3%)
1
( Fi R(s) Y(s)) + Td s( Fd R(s) Y(s))
Ti s
1
) E (s ) + K c ( b 1)R ( s) K cTd sY(s )
Tis
1
1 + 0.4Trs
U( s) = K c 1 +
+ Td s
R (s)
Y( s)
T s +1
Ti s
1 + sTr
1
1 + 0.4Tr s
U(s) = Kc 1 +
+ Td s
Y(s) K0 Y (s )
R (s )
Ti s
1 + sTr
Tf s + 1
Subtotal
32
3
(1%)
2
(1%)
1
(0%)
1
(0%)
61
(22%)
Total
124
42
18
17
49
30
1
Kc Td s
U(s) = Kc b +
Y(s) + Kc ( b 1) Y(s)
E( s)
Ti s
1 + Td s N
1
1 + Td s
U( s) = Kc 1 +
Y(s)
R( s)
Ts
Ti s
1+ d
1
U( s) = Kc 1 +
E( s) Kc Td sY ( s)
Ti s
1
U(s) = Kc b + [R (s) Y(s) ] ( c + Td s) Y(s)
Ts
i
U(s) =
Kc
E (s) K c (1 + Td s)Y(s)
Ti s
1
Td s
U(s ) = Kc 1 +
+
T
Ti s
1+ d
U(s) = K c Fp R(s) Y(s) +
U( s) = K c (1 +
3
(1%)
6
(2%)
280
3. Tuning rules for PI and PID controllers
Table 1: PI tuning rules - FOLPD model - G m ( s) =
K me s
1
ideal controller G c ( s) = Kc 1 +
. 84 tuning
1 + sTm
Ts
i
m
rules
Rule
Process reaction
Ziegler and Nichols [8]
Model: Method 1.
Hazebroek and Van der
Waerden [9]
Model: Method 1
m Tm
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
Kc
Ti
09
. Tm
K m m
3.33m
Tm
K m m
0.68
7.14
0.70
4.76
0.72
3.70
0.74
3.03
0.76
2.50
0.79
2.17
0.81
1.92
0.84
1.75
0.87
1.61
= 0.5 m + 01
.
Tm
m Tm
0.90
1.49
0.93
1.41
0.96
1.32
0.99
1.25
1.02
1.19
1.06
1.14
1.09
1.10
1.13
1.06
1.17
1.03
m
=
16
. m 1.2Tm
3.2m
Chien et al. [10] - regulator
0.6Tm
K m m
4 m
Chien et al. [10] - servo

Model: Method 1
Cohen and Coon [11]

process reaction
Model: Method 1.
07
. Tm
K m m
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
063
. Tm
Km m
Astrom and Hagglund [3]

regulator page 150
Model: Method 1
Quarter decay ratio.
m
1
Tm

page 138
Model: Not relevant
Model: Method 1
Comment
2.33
m
Km
m Tm
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
1.20
1.28
1.36
1.45
1.53
1.62
1.71
1.81
1.00
0.95
0.91
0.88
0.85
0.83
0.81
0.80
m
> 35
.
Tm
Ultimate cycle ZieglerNichols equivalent
0% overshoot
011
. < m < 10
.
Tm
20% overshoot
011
. < m < 10
.
Tm
07
. Tm
K m m
2.3m
20% overshoot
035
. Tm
Km m
117
. Tm
0% overshoot
011
. < m < 10
.
Tm
0.6Tm
K m m
Tm
20% overshoot
011
. < m < 10
.
Tm
1 Tm
+ 0083
.
0.9
K m m
3.33 m + 0.31 m
Tm
Tm
Tm
1 + 2.22 m
Tm
Quarter decay ratio
Rule
Kc
Ti
m Tm
Tm
K m m
Two constraints criterion Wolfe [12]
m Tm
0.2
0.5
4.4
1.8
3.23
2.27
1.0
5.0
0.78
0.30
1.28
0.53
Model: Method 2.
Two constraints criterion Murrill [13] page 356
Tm
0928
.

K m m
0.946
Tm m

1078
.
Tm
Comment
Decay ratio = 0.4; minimum
error integral (regulator
mode).
0 .583
Quarter decay ratio;

minimum error integral
(servo mode).
01
. m 10
.
Tm
Model: Method 3
McMillan [14] page 25
Model: Method 3
St. Clair [15] page 22
Model: Method 3
Shinskey [15a]
Model: Method 1
Regulator tuning
Minimum IAE - Murrill [13]
pages 358-363
Model: Method 3
Time delay dominant

processes
Tm
Tm
30
.
m
3.78 m
Tm
= 0.167
m
Km
3
0333
. Tm
K m m
0. 667 Tm
K m m
Performance index minimisation
Tm
0984
.

Km m
0.986
Tm m

0608
.
Tm
0. 707
01
.
100
. Tm Km m
30
. m
m Tm = 0.2
Minimum IAE - Shinskey

[16] page 123
Model: Method 6
104
. Tm K m m
2.25 m
m Tm = 0.5
111
. Tm Km m
145
. m
m Tm = 1
139
. Tm K m m
m Tm = 2

[17] page 38
Model: Method 6
Minimum IAE Huang et
al. [18]
Model: Method 6
0.95Tm K m m
34
. m
m Tm = 01
.
0.95Tm K m m
2.9 m
m Tm = 0.2
Minimum IAE Yu [19]

K e s L
(Load model = L
)
1 + sTL
Model: Method 6
Kc
0.685 TL
K m Tm
( 1) 1
Ti
m
0.214 0 .346
Tm
Tm
1.256
Tm TL

0. 214 Tm
01
.
( 1)
m
055
.
Tm
1977
.
m

Tm
1123
.
TL
Tm
m
Tm
TL
0874
.

Km Tm
0099
.
+ 0159
.
m
Tm
m

Tm
1041
.
Tm TL

0.415 Tm
4 .515
m
+ 0. 067
Tm
m

Tm
0 .876
2. 641
Kc
( 1)
1
=
Km
m
1
Tm
2 .641
m
Tm
m
Tm
+ 016
. ;
0.35
TL
+ 0.16
m
Tm
m
10
.
Tm
Tm
1;
0.35
0.9077
0.063
0.5961

m
m
6.4884 + 4.6198 m + 0.8196 m

52132
.
7.2712
0.7241e T

Tm
Tm
Tm
Tm
2
3
4
5
6
m
m
m
m
m
Ti (1) = Tm 00064
.
+ 3.9574 m 64789
.
+
9
.
4348
10
.
7619
+
7
.
5146
2
.
2236

Tm
Tm
Tm
Tm
Tm
Tm
Kc
Rule
Minimum IAE Yu [19]
K e s L
(Load model = L
)
1 + sTL
(continued)
Model: Method 6
Minimum ISE - Hazebroek
and Van der Waerden [9]
Ti
0015
. + 0 .384 m
Tm
TL
0871
.

K m Tm
TL
0513
.

K m Tm
0 .218
m
Tm
m

Tm
m

Tm
1055
.
1451
.
Tm TL

0. 444 Tm
Comment
0. 217 m 0. 213
Tm
Tm TL

0. 670 Tm
Tm

. + 0.3 m
074
K m m
Tm
0. 003
m
0. 084
Tm
m

Tm
0867
.
m

Tm
0 .56
m Tm
m Tm
0.2
0.3
0.5
0.80
0.83
0.89
7.14
5.00
3.23
0.7
1.0
1.5
0.96
1.07
1.26
2.44
1.85
1.41
2.0
3.0
5.0
0.959
Tm m

0492
.
Tm
0.739
0.945
Tm m

0535
.
Tm
0.586
Minimum ISE Zhuang and

Atherton [20]
Tm
1279
.

K m m
Tm
1346
.

Km m
0 .675
Tm m

0552
.
Tm
0. 438
m
Tm
0 .181 0 .205
m
Tm
0. 045 + 0. 344
TL
1289
.

K m Tm
Minimum ITAE - Murrill

[13] pages 358-363
Model: Method 3
Minimum ITAE Yu [19]
K e s L
(Load model = L
)
1 + sTL
Model: Method 6
0. 04 + 0. 067
m
Tm
m
Tm
Tm
0859
.

Km m
TL
0598
.

K m Tm
1. 214
m

Tm
Tm TL

0. 430 Tm
m
0 .49
Tm
0. 954
m

Tm
0 .639
m

Tm
1. 014
Tm TL

0. 359 Tm
m
0 .292
Tm
2. 532
m
10
.
Tm
11
.
m
2.0
Tm
TL
2.310
Tm
m

Tm
0. 899
1047
.
m

Tm
0889
.
0. 977
m
Tm
0. 272 0. 254
m

Tm
m

Tm
Tm TL

0.347 Tm
m
0. 094
Tm
1112
.
Tm TL
0596
.
Tm
m
0. 44
Tm
0 .372
Tm m

0674
.
Tm
1341
.
Tm TL

0805
.
Tm
0 .304
Tm
2 .310
m
Tm
m

Tm
m

Tm
Tm
Tm
01
.
0. 196
TL
Tm
m
Tm
0 .011 1945
.
m

Tm
1. 055
Tm TL

0. 425 Tm
5. 809
m
+ 0 .241
Tm
m

Tm
2 .385
m
Tm
0 .084 + 0154
.
m
Tm
m

Tm
1. 042
Tm TL

0. 431 Tm
0148
.
m
0. 365
Tm
m

Tm
0. 901
1<
TL
Tm
TL
Tm
1;
0.35
m
Tm
0.35
> 035
.
m
10
.
Tm
m
Tm
+ 0.112 ;
0.35
+ 0112
.
Tm
0.787 TL

Km Tm
+ 0. 077 ;
0.35
2.385
m
0 .901
Tm
3;
m
0. 46
0.680
m
0. 112
Tm
TL
1<
+ 0.077
m
0 .898
01
.
Tm
0. 735 TL

K m Tm
> 035
.
1.46
1.18
1.89
0.95
2.75
0.81
01
. m 10
.
Tm
Tm
0 .065+ 0 .234
0.35
Tm
107
. TL

K m Tm
Tm
TL
1157
.

Km Tm
m Tm < 0.2
Tm
1305
.

K m m
Model: Method 6
Tm
m Tm
0. 921 TL

K m Tm
3;
m
143
. Tm
Minimum ISE - Murrill [13]

pages 358-363
Model: Method 3
Minimum ISE Yu [19]

K e s L
(Load model = L
)
1 + sTL
Tm
Tm
K m m
Model: Method 1
Model: Method 6
TL
1<
TL
Tm
1;
0.35
3;
m
Tm
0.35
Kc
Rule
Minimum ITAE Yu [19]
K e s L
(Load model = L
)
1 + sTL
TL
0878
.

K m Tm
0172
.
0. 057
Ti
m
Tm
Tm
0 .909
Tm TL

0. 794 Tm
0 .228
Comment
m
0257
.
Tm
m

Tm
0. 489
> 035
.
Tm
(continued)
Model: Method 6
Minimum ISTSE - Zhuang
and Atherton [20]
Model: Method 6
Minimum ISTES Zhuang

and Atherton [20]
Model: Method 6
Servo tuning
Minimum IAE Rovira et al.
[21]
Model: Method 3
Minimum IAE - Huang et al.
[18]
Model: Method 6
Minimum ISE - Zhuang and
Atherton [20]
Model: Method 6
Minimum ISE Khan and

Lehman [22]
Model: Method 6
Minimum ITAE Rovira et

al. [21]
Model: Method 3
Tm
1015
.

K m m
0.957
Tm m

0667
.
Tm
0.552
Tm
1065
.

K m m
0.673
Tm m

0687
.
Tm
0.427
Tm
1021
.

Km m
0.953
Tm m

0629
.
Tm
0.546
Tm
1076
.

Km m
0 .648
Tm m

0650
.
Tm
0. 442
Tm
0758
.

K m m
0.861
Kc( 2) 2

Tm
1.020 0.323 m
Tm
Ti (2 )
Tm
0980
.

Km m
0.892
Tm
1072
.

K m m
0.560
m
2.0
Tm
01
.
m
10
.
Tm
11
.
m
2.0
Tm
01
.
m
10
.
Tm
01
.
m
1
Tm
m
10
.
Tm
Tm
11
.
m
2.0
Tm
Kc Km
0.01
m
0.2
Tm
0.2
m
20
Tm
01
.
m
10
.
Tm
0.648 0.114 m
Tm
0 .916
11
.
01
.
0.5291 00003082
2
m
Tm m
Kc Km
0.808 0.511 0.255 T
m
+
m
Tm
Tm m K m 0.095 + 0.846 0.381
2
m Tm m Tm m
m
Tm
0586
.

Km m
m
10
.
Tm
Tm
0.690 0.155 m
Tm
0.7388 03185
Tm
.
+
Tm Km
m
01
.
Tm
m
1.030 0165
.
Tm
1.0169
3.5959
3.6843
m
m
m
1
m
T
130454
Kc =
.
9.0916
+ 0.3053
+ 11075
.
2.2927
+ 4.8259e

Km
Tm
Tm
Tm
Tm
2
3
4
5
6
m
m
m
m
m
m
(2 )
Ti = Tm 0.9771 0.2492
+ 3.4651 7.4538 + 8.2567 4.7536 + 11496
.

Tm
Tm
Tm
Tm
Tm
Tm
( 2)
Rule
Minimum ISTSE - Zhuang
and Atherton [20]
Model: Method 6
Minimum ISTES Zhuang

and Atherton [20]
Model: Method 6
Direct synthesis
Haalman [23]
Model: Method 6
Chen and Yang [23a]
Model: Method 25
Minimum IAE regulator Pemberton [24], Smith and
Corripio [25] page 343346. Model: Method 6
Minimum IAE servo Smith and Corripio [25]
page 343-346. Model:
Method 6
5% overshoot servo
Smith et al. [26] deduced
from graph.
Model: Method not stated
1% overshoot servo
Smith et al. [26] deduced
from graph
5% overshoot - servo Smith and Corripio [25]
page 343-346. Model:
Method 6
5% overshoot - servo Hang et al. [27]
Model: Method 1
Miluse et al. [27a]
Kc
Ti
Tm
0712
.

Km m
0.921
Tm
0786
.

Km m
0.559
Tm
0569
.

K m m
0.951
Tm
0628
.

Km m
0.583
2Tm
3K m m
0. 7T m
Km m
Comment
Tm
01
.
m
10
.
Tm
Tm
11
.
m
2.0
Tm
Tm
01
.
m
10
.
Tm
Tm
11
.
m
2.0
Tm
0.968 0.247 m
Tm
0.883 0.158 m
Tm
1.023 0.179 m
Tm
1.007 0.167 m
Tm
Tm
Closed loop sensitivity

M s = 19
. . (Astrom and
Hagglund [3])
Ms = 1. 26 ; A m = 2. 24 ;
Tm
m = 500
01
.
m
0.5
Tm
Tm
01
.
m
05
.
Tm
052
. Tm
K m m
Tm
0.04
m
14
.
Tm
044T
. m
K m m
Tm
0.04
m
14
.
Tm
Tm
Km m
3Tm
5Km m
Tm
Tm
2K m m
Tm
13Tm
25K m m
Tm
0. 368Tm
K m m
Tm
0. 514Tm
K m m
Tm
Closed loop response

overshoot = 0%
(Model: Method 26 Miluse
et al. [27b])
overshoot = 5%
0. 581Tm
K m m
Tm

overshoot = 10%
Rule
Kc
Ti
Comment
Miluse et al. [27a] continued
0. 641Tm
K m m
Tm

overshoot = 15%
0. 696Tm
K m m
Tm

overshoot = 20%
0. 748Tm
K m m
Tm

overshoot = 25%
0. 801Tm
K m m
Tm

overshoot = 30%
0. 853Tm
K m m
Tm

overshoot = 35%
0. 906Tm
K m m
Tm

overshoot = 40%
0. 957 Tm
K m m
Tm

overshoot = 45%
1.008 Tm
Km m
Tm

overshoot = 50%
Regulator Gorecki et al.

[28]
(considered as 2 rules)
Model: Method 6
Chiu et al. [29]
Model: Method 6
Kc
Astrom et al. [30]
Kc
Ti
Ti ( 4)
Low freq. part of magnitude

Bode diagram is flat
Tm
K m (1 + m )
Tm
variable; suggested
values: 0.2, 0.4, 0.6, 1.0.
Model: Method 12
K m 1.905 m + 0.826
Tm

2 Tm
=
2 + m 1 e
K m m
2Tm
= m
Tm
Honeywell UDC 6000

controller
m
.
+ 0.362 Tm
0153
Tm

damping factor =
0.9; 02
. m Tm 1 .
m
m
2 +
2
2Tm
2 Tm

1+ m
2Tm
2
3
2
2

3 + m + m + m 2 + m 2 + m
2Tm Tm
2Tm
2Tm
2Tm
2
Kc ( 4) =
Pole is real and has max.

attainable multiplicity
( 3)
Kc( 4) 3
Davydov et al. [31]
( 3)
( 3)
Ti
3
3 m
K m1 +
Tm
Model: Method 22
( 3)
1
Km
T
T
Tm
+ 6 m + 6 m
m
m
m
2
Tm
Tm
4 1+ 3
+ 3
m
m
1+3
T
T
Tm
+ 6 m + 6 m
m
m
m
2
Tm
Tm
3 1+ 2
+ 2
m
m
1+ 3
, Ti ( 4) = m
Kc
Rule
Ti
Comment
Schneider [32]
0.368
Tm
K m m
Tm

damping factor = 1
Model: Method 6
0.403
Tm
K m m
Tm

damping factor = 0.6
McAnany [33]
(1.44Tm + 0.72Tm m 0.43 m 2.14)
K m 12
. m + 0.36 m 2 + 2
Model: Method 5
Leva et al. [34]
cn Tm
Km
Model: Method 16
Khan and Lehman [22]
Tm
Km
Tm
Km
Model: Method 11
Or
Model: Method 14
Gain and phase margin Ho

et al. [37]
Closed loop time constant =

167
. Tm .
tan m + m cn + tan 1 ( cn Tm )
2
cn
1 + cn 2 Tm 2
1 + cn 2 Ti 2
0.404 0.256
01275
.
m
T
m Tm
cn
Tm
2.82 m tan 1
m

m 2
=
m
KcKm
0.4104 0.00024 0.525
m2
Tm 2
m Tm
KcKm
0.719 0.0808
0.324
2
m Tm
m mTm
0.01
m
0.2
Tm
0.2
m
20
Tm
1048
. Tm
Km m
Tm
Gain Margin = 1.5

Phase Margin = 30 0
07854
.
Tm
K m m
Tm
Gain Margin = 2
Phase Margin = 45 0
0524
. Tm
K m m
Tm
Gain Margin = 3
Phase Margin = 60 0
0.393Tm
K m m
Tm
Gain Margin = 4
Phase Margin = 67 .5 0
0314T
.
m
Km m
Tm
Gain Margin = 5
Phase Margin = 72 0
p Tm
K m Am
4 p m
p =
2 p
(considered as 2 rules)
Model: Method 6
4Tm + 128
. m 2.4
0. 3852 0.723

+
0.404 m2
Tm
m
Tm
Model: Method 6
Hang et al. [35, 36]
Km 556
. + 2 m + m
A m m + 0.5A m (A m 1)
(A
2
m
1 m
Tm
Given A m , ISE is minimised when m = 688884

.
34.3534 A m + 91606
.
m
for servo
Tm
tuning (Ho et al [104]).

Given A m , ISE is minimised when m = 45.9848A m 0.2677 ( m Tm )
0. 2755
for regulator
tuning (Ho et al [104]).
Tan et al. [39]
Ti 1 + Tm
Model: Method 10
A m 1 + Ti
Symmetrical optimum
principle - Voda and Landau
[40]
< u
4.6
135
35
. G p ( j135 )
0
1
Model: Method not relevant
tan tan 1 T m m
2.828 G p ( j135 )
1
115
. G p ( j 135 0 ) + 0.75K m
4.6 G p ( j135 ) 0.6K m

0
= 08
. ,
m
< 05
. ;
Tm
= 0.5,
m
> 05
.
Tm
m
01
.
Tm
4
135
01
. <
0
135 0 2.3 G p ( j 1350 ) 0.3K m
m
015
.
Tm
0.15 <
m
1
Tm
Rule
Kc
Ti
Comment
Friman and Waller [41]

Model: Method 6
0.2333
m > 2Tm . Gain margin = 3;
G p ( j135 )
135
Model: Method 6
Tm
2K m m
Tm
Smith [42]
Model: Method not
specified
0.35
Km
042
. m
Kc( 5)
Ti (5)
0.5Tm
Km m
Tm
Voda and Landau [40]
Modulus optimum principle

- Cox et al. [43]4
Model: Method 17
Cluett and Wang [44]

Model: Method 6
Bi et al. [46]
Model: Method 20
Kc
( 5)
0.5
Km
6 dB gain margin - dominant

delay process
m
1
Tm
m
>1
Tm
0.099508 m + 0.99956Tm
m
4m
0.99747 m 8742510
.
. 5 Tm
0.05548 m + 0.33639Tm
K mm
016440
.
m + 0.99558Tm
m
2m
0.98607 m 15032
.
.10 4 Tm
0.092654 m + 0.43620Tm
K mm
0.20926 m + 098518
.
Tm
m
133
. m
0.96515 m + 4.255010
. 3 Tm
012786
.
m + 051235
.
Tm
Km m
0.24145 m + 0.96751Tm
0.93566 m + 2.298810
. 2 Tm m
m
016051
.
m + 0.57109Tm
Km m
0.26502 m + 0.94291Tm
089868
.
m + 6.935510
. 2 Tm m
08
. m
0.28242 m + 0.91231Tm
0.85491 m + 015937
.
Tm m
1.045
Model: Method 6
Phase margin = 60 0 ;
0.25 m 1
Tm
0.019952 m + 0.20042Tm
K mm
019067
.
m + 0.61593Tm
Km m
Abbas [45]
Phase margin = 45 0

0148
.
+ 0186
. m
Tm
Km (0.497 0.464 V0.590 )
05064
.
Tm
Km m
Tm + 0.5 m

067
. m
V = fractional overshoot,
0 V 0.2
01
. m 5.0
Tm
Tm
Tm 3 + Tm2 m + 0.5Tm m 2 + 0.167 m 3

Tm3 + Tm 2 m + 0.5Tm m 2 + 0.167 m 3
( 5)
T
=
i
2
2
3
2
2
Tm m + Tm m + 0.667 m
Tm + Tm m + 0.5 m
Kc
Rule
Wang and Shao [47]
Kc
Ti
( 5a ) 5
Ti
Comment
= inverse of the maximum

of the absolute real part of
the open loop transfer
function; = [1.5, 2.5]
(5 a )
Model: Method 6
Robust
Tm + 0.5 m
K m( + m )
Brambilla et al. [48] -
Closed loop response has less than 5% overshoot with no model uncertainty:
= 1 , 01
. m 1 ; = 1 05
. log 10 m , 1 < m 10
Tm
Tm
Tm
Model: Method 6
Rivera et al. [49]
Tm
Km
Tm
17
. m , > 01
. Tm .
Model: Method 6
2Tm + m
2K m
Tm + 0.5 m
17
. m , > 01
. Tm .
Tm
K m ( m + )
Tm
m
2Km ( m + )
05
. m
Fruehauf et al. [52]
5Tm
9 m Km
5 m
m
< 0.33
Tm
Model: Method 1
Tm
2 m Km
Tm
m
0.33
Tm
Chen et al. [53]
0. 50Tm
m K m
Tm
A m = 3.14 , m = 61. 40 ,
Ms = 1. 00
Tm
A m = 2.58 , m = 55. 00 ,
Ms = 1. 10
0. 67Tm
mK m
Tm
A m = 2. 34 , m = 51.6 0 ,
Ms = 1. 20
0. 70Tm
m K m
Tm
A m = 2. 24 , m = 50.0 0 ,
M s = 1. 26
0. 72Tm
m K m
Tm
A m = 2.18 , m = 48. 70 ,
Ms = 1. 30
Chien [50]
Model: Method 6
Thomasson [51]
Model: Method not defined
0. 61Tm
m Km
Model: Method 6
Kc
( 5a )
f 1 (90
( 5a )
= Tm [50];
> Tm + m , m << Tm
(Thomasson [51])
m >> Tm ; = desired
closed loop time constant
1
1
f 2 ( 900 )
,
f 1 (90 0 )
90 0
Km
2
2
2
)=
( Tm + {1 + 90 Tm } m ) sin(90 m tan 1 90 Tm ) 90 Tm cos( 90 m tan 1 90 Tm )
2
2 1 .5
1 + 90 Tm
=
f 2 ( 90 ) =
Ti
Tm + 0.5 m
1
1 + 90 Tm
2
[(T
+ {1 + 90 Tm } m ) cot(90 m tan 1 90 Tm ) + 90 Tm
2
900 Tm + (1 + 90 0 2Tm 2 ) m cos( 90 0 m tan 1 90 0 Tm ) + (1 + 2 900 2 Tm 2 ) sin( 90 0 m tan 1 900 Tm )

900 Tm cos(90 0 m tan 1 90 0 Tm ) + 90 0 [Tm + (1 + 900 Tm ) m ] sin( 900 m tan 1 90 0 Tm )
3
Rule
Kc
Chen et al. [53]

- continued
Model: Method 6
Ogawa [54] deduced from

graphs
Model: Method 6
Tm
A m = 2. 07 , m = 46. 50 ,
Ms = 1. 40
0. 80Tm
m K m
Tm
A m = 1.96 , m = 44.10 ,
Ms = 1. 50
Tm
m Tm
0.5
1.0
0.5
1.0
0.5
1.0
0.5
1.0
0.9
0.6
0.7
0.47
0.47
0.36
0.4
0.33
1.3
1.6
1.3
1.7
1.3
1.8
1.3
1.8
2.0
10.0
2.0
10.0
2.0
10.0
2.0
10.0
0.45
0.4
0.4
0.35
0.32
0.3
0.3
0.29
2.0
7.0
2.2
7.5
2.4
8.5
2.4
9.0
Ti
K m ( + m )
Tm +
Km ( m + )
Model: Method 21
Ultimate cycle
m
2( + m )
2
= 0.333 m
Tm > m
Tm ( m + 2 ) 2
Tm ( m + 2 ) 2
Chun et al. [57]
20% uncertainty in the

process parameters
process parameters
process parameters
process parameters
Desired closed loop
e m s
response =
,
( s + 1)
Tm + 025
. m
Tm + 0.25 m
Km
Isaksson and Graebe [56]

Model: Method 6
= 0.4Tm
m + Tm
0.65
T

1881
.
Tm
1
m
166
. m 1 +

0.65
Km m T
Tm + m
m
1 + T +
m
m
Regulator minimum IAE

Shinskey [59] page 167.
Model: Method not
specified
Model: Method not
specified
T
3.05 0.35 u
m
1 =
Tuning rules developed

from Ku , Tu
T
T
Tu 0.87 0.855 u + 0.172 u
m
m
Ku

Shinskey [16] page 148
Model: Method 6
Regulator nearly minimum
IAE, ISE, ITAE Hwang
[60]
Model: Method 8
0. 76Tm
m K m
m Tm
Model: Method 6
Model: Method 1 or
Method 18
Comment
Km
Lee et al. [55]
McMillan [58]
Ti
055
. Ku
0.78Tu
m Tm = 0.2
0. 48Ku
0.47 Tu
m Tm = 1
0.5848Ku
0.81Tu
m Tm = 0.2
0.5405K u
0.66Tu
m Tm = 0.5
0.4762K u
0.47Tu
m Tm = 1
0.4608 K u
0.37Tu
m Tm = 2
( 1 1 ) K u ,
114
. 1 0.482 u m + 0.068 2u 2m
Ku K m 1 + K u K m
Kc 2 Ku u ,
2 =
0.0694 1 + 2.1 u m
0.367 2u m2
Ku Km 1 + Ku Km
Decay ratio = 0.15

01
. m 2.0
Tm
Kc
Rule
[60]
(continued)
Model: Method 8
Servo small IAE Hwang

[60]
1 =
Model: Method not

specified
0.0724 2u 2m
K u K m 1 + K u Km
( 1 1 ) K u ,
1 =
1 =
1 =
1 =
Hang et al. [65]
1.09 1 0.497 u m +
Model: Method 8
r = parameter related to the
position of the dominant real
pole.
( 1 1 ) K u ,
1.03 1 0.51u m +
0.0759 u2 m2
Ku Km 1 + Ku Km
1.07 1 0.466 u m + 0.0667 u2 m2

Ku Km 1 + Ku Km
1 + 2.54 u m 0.457u2 m2
K u Km 1 + K u Km
( 1 1 ) K u ,
0.0647 2u m2
0.0386 1 + 3.26 u m 0.6 2u m2

K u Km 1 + K u Km
2 =
0.0328 1 + 2.21 u m
Decay ratio = 0.25

01
. m 2.0
Tm
0.338 2u m2
K u Km 1 + K u Km
2 =
0.0477 1 + 2.07 u m
0.333 2u 2m
2 =
Ku K m 1 + K u K m
Kc 2 Ku u ,
Decay ratio = 0.1, r = 0.5

01
. m 2.0
Tm
Kc 2 Ku u ,
0.0657 2u 2m
Ku Km 1 + Ku Km
Decay ratio = 0.2

01
. m 2.0
Tm
0.054
2 =
2 =
Ku Km 1 + K u K m
114
. 1 0.466u m +
Kc 2 Ku u ,
Kc 2 Ku u ,
( 1 1 ) K u ,
Comment
Kc 2 Ku u ,
( 1 1 ) K u ,
111
. 1 0.467 u m +
Ti
0.0609 1 + 197
. u m 0.323u2 m2
11 T + 13
12 + 2
m
37 m 4
4
0.2
5
15
Tm
Ku
6
m
11 T + 13
m
15 + 14
37
4
Tm
Ku Km 1 + Ku Km

01
. m 2.0
Tm
01
. m 2.0
Tm
11 T + 13
016
. m < 096
. ;
m
+ 1Tu
Tm
37 m 4
Servo response: 10%
T

m

overshoot, 3% undershoot
Servo minimum ISTSE

Zhuang and Atherton [20]
0.361K u
0.083 (1.935 K m K u + 1)Tu
Regulator minimum ISTSE 1892
. K m K u + 0244
.
0.706Km K u 0.227
Ku
Ku
Zhuang and Atherton [20]
3249
.
K
K
+
2
.
097
.
0.7229Km K u + 12736
m u
2
( Kc / K u u )
Servo nearly minimum IAE
m
m
2
.
+ 0.0376
Ku
m
and ITAE Hwang and 0.438 0110
m
Tm
T
m
0.0388 + 0108
.
0.0154

Tm
T m
Fang [61]
Model: Method 9
2
( Kc / K u u )
m
m
2
m
IAE and ITAE Hwang and 0.515 0.0521 Tm + 0.0254 Tm Ku
m
0.0877 + 0.0918
0.0141
Tm
Tm
Fang [61]
Model: Method 9
2
( Kc / K u u )
Simultaneous

0.46 0.0835 m + 0.0305 m K
2
u

Servo/regulator Hwang
Tm
Tm
0.0644 + 0 .0759 m 0.0111 m
Tm
and Fang [61]
Tm
Model: Method 9
01
.
m
2.0
Tm
01
.
m
2.0
Tm
m
2.0 ;
Tm
decay ratio = 0.03
01
.
m
2.0 ;
Tm
decay ratio = 0.12
01
.
01
.
m
2.0
Tm
K me s
1
Controller G c ( s) = Kc b +
. 2 tuning rules
1 + sTm
Ti s
Kc
Rule
Ti
Direct synthesis
(Maximum sensitivity)
029
. e
Astrom and Hagglund [3] dominant pole design
page 205-208
Comment
2.7 + 3.7 2
Tm
K m m
8.9 m e 6.6 + 3.0 or

2
= m ( m + Tm )
0.79Tm e 1.4 + 2.4
Model: Method 3 or 4
078
. e 4.1 + 5.7 Tm
Km m
8.9 m e 6.6 + 3.0 or
Astrom and Hagglund [3] modified Ziegler-Nichols

page 208
04
. Tm
K m m
0.7Tm
0.79Tm e 1.4 + 2.4
b = 0.81e0.73 + 1.9
Ms = 1.4 , 014
. m 55
.
Tm
2
b = 0.44e0.78 0.45 ;
Ms = 2.0, 014
. m 55
.
Tm
2
b = 0.5; 01
.
m
2
Tm
K me s
Two degree of freedom controller:
1 + sTm
m
1
E (s ) Kc R( s) . 1 tuning rule
U(s) = K c 1 +
Tis
Rule
Minimum servo/regulator
Taguchi and Araki [61a]
Model: ideal process
Kc
Ti
1
0.7382
0
.
1098
+
m
K m
0.002434
Tm
Ti
( 5b ) 6
= 0. 6830 0.4242 m + 0.06568 m

Tm
Tm
Ti
( 5b )
Comment
2
3
= Tm 0.06216 + 3.171 m 3.058 m + 1. 205 m
Tm
Tm
Tm
m
1.0 .
Tm
Overshoot (servo step)
20% ; settling time
settling time of tuning
rules of Chien et al. [10]

1
Table 4: PI tuning rules - non-model specific controller G c ( s) = Kc 1 +
. 20 tuning rules
Ts
i
Kc
Ti
Comment
0.45Ku
0.83Tu
Quarter decay ratio
0.45Ku
1
5.22
.
522
p1
T1
p1 , T1 = decay rate, period

measured under
proportional control when
Kc = 05
. Ku
** Hang et al. [36]
0.25Ku
0.2546Tu
dominant time delay process
McMillan [14] page 90
0.3571K u
Tu
Pessen [63]
0.25Ku
0.042 K uTu
dominant time delay process
0.4698 K u
0.4373Tu
01988
.
Ku
0.0882Tu
0.2015 Ku
01537
.
Tu
Parr [64] page 191
05
. Ku
043
. Tu
Gain margin = 2, phase

margin = 20 0
Gain margin = 2.44, phase
margin = 61 0
Gain margin = 3.45, phase
margin = 46 0
Quarter decay ratio
Yu [122] page 11
0.33Ku
2Tu
Other tuning rules

Parr [64] page 191
0667
. K25%
T25%
Quarter decay ratio
042
. K25%
T25%
Fast tuning
0.33K25%
T25%
Slow tuning
0333
.
G p ( j u )
2Tu
Bang-bang oscillation test
0.5
Alfa-Laval Automation
ECA400 controller
Rule
Ultimate cycle
Hwang and Chang [62]

page 142
McMillan [14] pages 4243

Parr [64] page 192
Hagglund and Astrom [66]
G p ( j 135 )
135
0.25
1.6
135
G p ( j 135 )
0
Alfa-Laval Automation
ECA400 controller - process
has a long delay
Kc
Rule
tan m p 0 .5
Leva [67]
Ti
tan m p 05
.
G p ( j ) 1 + tan m p 05
.
sin m
tan m
90
Astrom [68]
G p ( j 90 )
Comment
m > p + 0.5 ,
p = process phase at
frequency
1
u
m = 450 ,small m
016
. Tu tan m
V = relay amplitude, A = limit

cycle amplitude.
A3
A2
A m 2 , m 60 0
1
Calcev and Gorez [69]
2 2 G p ( j u )
Cox et al. [70]
020
. VTu sin m
A
m = 15 0 , large m
Direct synthesis
Vrancic et al. [71]
05
. A3
A1A2 Km A 3
Vrancic [72]
05
. A3
A1A2 Km A 3
333
. m
3.7321
150
Gain margin = 2,
Phase margin = 30 0
1.18K u K m 1.72
(0.33K u K m 0.17) u
0.1 Ku Km 0.5
0. 50K u K m 0. 36
(0.33K u K m 0.17) u
K u K m > 0.5
0.4830
Friman and Waller [41]
G p ( j150 )
1.18 K u K m 1. 72
Kristiansson and
Lennartson [158a]
K uK m
0. 50K u K m 0. 36
Ku Km
20K135 Km 160
20K135 K m 160
K135 K m
5.4K135 K m 13. 6
0
K135 K m
(0.315 K135 Km 0.175 ) u

5. 4K135 K m 13. 6
K u K m < 0. 1 ;
K 135 K m > 0 .1
(1.32K135 K m 3.2) u
0
A1 = y1( ) , A 2 = y 2 ( ) , A 3 = y3 ( )
Alternatively, if the process model is G m (s) = K m
1 + b1 s + b2 s2 + b 3s3 s
e
, then
1 + a1s + a 2 s3 + a 3s3
m
A1 = Km ( a1 b1 + m ) , A2 = K m b 2 a2 + A1a1 b1 m + 05
. m2 ,
K u K m < 0. 1 ;
K 135 0 K m 0 .1
Modified Ziegler-Nichols
process reaction method
A3 = Km a 3 b3 + A 2 a1 A1a 2 + b2 m 05
. b1 m 2 + 0167
. m3

1
Table 5: PI tuning rules - non-model specific controller G c ( s) = K c b +
. 1 tuning rule
Ti s
Rule
Kc
Ti
Direct synthesis
2.9 2.6 2
Astrom and Hagglund [3] Ms = 1.4 page 215
Comment
0.053 Ku e
= 1 Km Ku
b = 11
. e 0.0061 + 1.8 ;
0 < Km K u < .
maximum sensitivity Ms
=1.4
2
0.90Tu e 4.4 + 2. 7
b = 048
. e 0.40 0.17 ;
0 < Km K u < .
Ms = 2.0
2
013
. K ue
1.9 1.3 2
0.90Tu e
4.4 + 2. 7 2

1
Table 6: PI tuning rules - non-model specific controller U( s) = Kc 1 +
E( s) + Kc ( b 1)R (s) . 1 tuning rule
Ti s
Kc
Rule
Ti
Comment
Direct synthesis
b = [0.5,08
. ] - good servo
A1
Vrancic [72]
Kc
Kc
( 6)
( 6)
A 1A 2 K m A 3
A 1A 2 K m A 3 +
(1 b )( K
2 K c (6 )
K c( 6) K m 2
2
(1 b )
2
A 3 + A 1 2K mA 1A 2
3
( K m A3 A 1A 2 ) 2 (1 b2 )A 3 ( Km 2 A3 + A13 2K m A1A2 )
(1 b )( K
y( )
y1( t) = Km
d , y2 ( t) =
Km +
and regulator response
( K mA 3 A 1A 2 ) 2 (1 b2 )A 3 ( Km 2 A3 + A13 2K mA 1A2 )
Kc
( 6)
A 3 + A 1 2K mA 1A 2
3
(A 1 y1 ())d ,
0
y3 ( t) =
(A
0
y2 ( ) )d
, Km A 3 A1A 2 < 0
, Km A 3 A1A 2 > 0
Table 7: PI tuning rules - non-model specific controller U( s) = Kc Y(s)
Kc
E ( s) . 1 tuning rule
Ti s
Kc
Ti
Comment
TCL + 1414
.
TC LTm + mTm
T CL + 1414
. T C L Tm + Tm m
Underdamped system
response - = 0.707 .
Rule
Direct synthesis
2
Chien et al. [74]
K m TC L2
+ 1414
. TCL m + m
Tm + m
m > 0.2Tm
Table 8: PI tuning rules - IPD model G m ( s) =
K me s m
s
- controller G c ( s) = Kc 1 + 1 . 21 tuning rules
Rule
Kc
Ti
Process reaction
Model: Method 1.
0.9
Km m
333
. m
0.6
Km m
2.78 m
Two constraints method

Wolfe [12]
Model: Method 1
087
.
Km m
4.35m
0.487
Km m
8.75 m
Tyreus and Luyben [75]

page 138
Model: Not relevant
Regulator tuning
Minimum ISE Hazebroek
and Van der Waerden [9]
Model: Method 1
Shinskey [59] minimum
IAE regulator page 74.
Model: Method not
specified
Poulin and Pomerleau [82]
minimum ITAE (process
output step load
disturbance)
Model: Method 2
minimum ITAE (process
input step load disturbance)
Model: Method 2
Ultimate cycle
Tyreus and Luyben [75]
0.63
Km m
Comment
Quarter decay ratio
Decay ratio = 0.4; minimum
error integral (regulator
mode).
Decay ratio is as small as
possible; minimum error
integral (regulator mode).
Maximum closed loop log
modulus = 2dB ; closed loop
time constant = m 10
3.2m
15
.
K m m
556
. m
09259
.
Km m
4 m
05264
.
K m m
4.5804 m
05327
.
K m m
38853
.
m
0.31Ku
2.2Tu
061
. Ku
Tu
05
.
Km m
5 m
Model: Method 2 or 3.
Model: method not
specified
Robust
Model: Method 5
Ts
i

modulus = 2dB ; closed loop
time constant = m 10
Rule
Kc
Chien [50]
1 2 + m
Km [ + m ]2
Model: Method 2
15
. m
Ogawa [54] deduced from

graph
Model: Method 5
Overshoot
58%
TS
6 m
Ti
Comment
2 + m
=
, m [50];
Km
> m + Tm (Thomasson
[51])
PP
17
. Km
m
RT
7Km
m
2.5 m
35%
11 m
2.0K m
m
16K m
m
35
. m
26%
16 m
2.2K m
m
23Km
m
45
. m
22%
20 m
25
. Km
m
30K m
m
Zhang et al. [135]

= [15
. m ,4.5 m ]
- values deduced from
graphs
11 m

process parameters
Kmm
12 m

process parameters
0.34
K m m
13 m

process parameters
0.30
Km m
14 m

process parameters
0.27
K m m
15 m

process parameters
0.45
K m m
0.39
Rule
Kc
Ti
K m m
Comment
Direct synthesis
Wang and Cluett [76]

deduced from graph
Model: Method 2

Model: Method 2
Rotach [77]
Model: Method 4
Model: Method 2
Closed Damp. Gain Phase

loop Factor margin margin
time
Am
m
const.
[deg.]
0.707
1.3
11
Closed Damp. Gain Phase

loop Factor margin margin
time
Am
m
const.
[deg.]
0.9056 2.6096 m
1.0
1.3
14 0.8859 3.212
2m
0.707
2.5
33
0.5501 4.0116
2m
1.0
2.3
37
0.6109 5.2005
3 m
0.707
3.6
42
0.3950 5.4136
3 m
1.0
3.0
46
0.4662 7.1890
4m
0.707
4.7
47
0.3081 6.8156
4m
1.0
4.0
52
0.3770 9.1775
5 m
0.707
5.9
50
0.2526 8.2176
5 m
1.0
4.8
56
0.3164 11.166
6 m
0.707
7.1
52
0.2140 9.6196
6 m
1.0
5.6
59
0.2726 13.155
7m
0.707
8.2
54
0.1856 11.022
7m
1.0
6.3
61
0.2394 15.143
8 m
0.707
9.2
55
0.1639 12.424
8 m
1.0
7.2
62
0.2135 17.132
9m
0.707
10.4
56
0.1467 13.826
9m
1.0
8.0
64
0.1926 19.120
10 m 0.707
11.5
57
0.1328 15.228 10 m
1.0
8.7
65
0.1754 21.109
11m 0.707
12.7
58
0.1213 16.630 11m
1.0
9.6
66
0.1611 23.097
12 m 0.707
13.8
59
0.1117 18.032 12 m
1.0
10.4
67
0.1489 25.086
13 m 0.707
14.9
59
0.1034 19.434 13 m
1.0
11.2
67
0.1384 27.074
14 m 0.707
16.0
60
0.0963 20.836 14 m
1.0
12.0
68
0.1293 29.063
15 m 0.707
17.0
60
0.0901 22.238 15 m
1.0
12.7
68
0.1213 31.051
16 m 0.707
18.2
60
0.0847 23.640 16 m
1.0
13.6
69
0.1143 33.040
0.9588
Km m
30425
.
m

m
0.6232
Km m
52586
.
m

2 m
0.4668
Km m
7.2291 m

3 m
0.3752
Km m
91925
.
m

4 m
0.3144
Km m
111637
.
m

5 m
0.2709
Km m
131416
.
m

6 m
0.75
K m m
241
. m
034
. K u or
2.13
Km Tu
104
. Tu
Damping factor for

oscillations to a disturbance
input = 0.75.
Maximum sensitivity
= 5 dB
Rule
Gain and phase margin
Kookos et al. [38]
Model: Method 2
Representative results
Kc
Ti
1
p 05
. p m
A m Km
Comment
p =
A m m + 0.5A m (A m 1)
(A
2
m
1 m
0.942
K m m
4510
. m
Gain Margin = 1.5

Phase Margin = 22 .5 0
0.698
K m m
4.098 m
Gain Margin = 2
Phase Margin = 30 0
0.491
K m m
6942
.
m
Gain Margin = 3
Phase Margin = 45 0
0384
.
K m m
18.710m
Gain Margin = 4
Phase Margin = 60 0
Other methods
Penner [79]
Model: Method 2
Srividya and Chidambaram

[80]
Model: Method 5
0.58
K m m
10 m
Maximum closed loop gain =

1.26
0.8
K m m
59
. m
Maximum closed loop gain =

2.0
0.67075
K m m
36547
.
m
Rule
K me s m
s
- controller G c ( s) = Kc 1 + 1
Kc
Ti
0463
. + 0.277
K m m
m
0.238 + 0.123
1
. 1 tuning rule
Ts
1
+
Tf s
i
Comment
Robust
Tan et al. [81]
Model: Method 2
Tf =
m
,
5.750 + 0.590
= 0.5
Kc
Rule
Direct synthesis
Chien et al. [74]
Model: Method 1
K me s m
s
1.414TCL + m
K m TCL + 1.414TCL m + m
2
- controller U( s) = Kc Y(s)
Kc
E ( s) . 1 tuning rule
Ti s
Ti
Comment
1414
. TCL + m
Underdamped system
m 0.2Tm
K me s m
s
- Two degree of freedom controller:
1
E (s ) Kc R( s) . 1 tuning rule
U(s) = K c 1 +
Tis
Rule
Kc
Servo/regulator tuning
Ti
0.049 Ku
2. 826 Tu
m
1.0 .
Tm
20% ; settling time
= 0.506 , c = 1640
0.066 Ku
2. 402 Tu
= 0.512 , c = 1600
0.099 Ku
1.962 Tu
= 0.522 , c = 1550
0.129 Ku
1.716 Tu
= 0.532 , c = 1500
0.159 Ku
1.506 Tu
= 0.544 , c = 1450
0.189 Ku
1.392 Tu
= 0.555 , c = 1400
0.218 K u
1.279 Tu
= 0.564 , c = 1350
0.250 Ku
1.216 Tu
= 0.573 , c = 1300
0.286 Ku
1.127 Tu
= 0.578 , c = 1250
0.330 Ku
1.114 Tu
= 0.579 , c = 1200
0.351K u
1.093Tu
= 0.577 , c = 1180
0. 7662
K m m
4.091 m
= 0. 6810
Minimum ITAE Pecharroman and Pagola

[134b]
Km =1
Model: Method 6
Comment
Table 12: PI tuning rules FOLIPD model G m (s) =
Km e s

s(1 + sTm )
Ts
i
m
Kc
Rule
Ti
Comment
Ultimate cycle
2
T 0.65
1477
.
Tm
1
332
. m 1 + m
m
Model: Method not relevant K m 2 0.65
from Ku , Tu
m
Tm
1
+
m
Regulator tuning
Minimum performance index
Minimum IAE Shinskey
[59] page 75.
0.556
3.7( m + Tm )
Model: Method not
K m ( m + Tm )
specified
[59] page 158
0.952
4(Tm + m )
Model: Open loop method
Km (Tm + m )
not specified
2
Minimum ITAE Poulin and
b
Tm
Pomerleau [82] deduced K m ( m + Tm ) a + T 2 + 1
a( m + Tm )
( m m)
from graph
McMillan [58]
Model: Method 2
m Tm
m Tm
m Tm
Output step load

disturbance
0.2
0.4
0.6
0.8
5.0728
4.9688
4.8983
4.8218
0.5231
0.5237
0.5241
0.5245
1.0
1.2
1.4
1.6
4.7839
4.7565
4.7293
4.7107
0.5249
0.5250
0.5252
0.5254
1.8
2.0
4.6837
4.6669
0.5256
0.5257
0.2
0.4
0.6
0.8
3.9465
3.9981
4.0397
4.0397
0.5320
0.5315
0.5311
0.5311
1.0
1.2
1.4
1.6
4.0397
4.0337
4.0278
4.0278
0.5311
0.5312
0.5312
0.5312
1.8
2.0
4.0218
4.0099
0.5313
0.5314
(2 tuning rules)
Input step load disturbance
Direct synthesis
Model: Method 2
034
. K u or
2.13
Km Tu
104
. Tu
Maximum sensitivity
= 5 dB
Table 13: PI tuning rules FOLIPD model G m (s) =
Rule
Kc
Km e s
- controller G c ( s) = Kc b + 1 . 1 tuning rule

s(1 + sTm )
Ti s
Ti
Comment
Direct synthesis
041
. e 0.23 + 0.019
,
Km ( Tm + m )
b = 0.33e2.5 1.9 .
Ms = 1.4; 014
. m 55
.
Tm
Astrom and Hagglund [3] maximum sensitivity pages

210-212
Model: Method 1.
57
. m e1.7 0.69
= m ( m + Tm )
081
. e 1.1 + 0.76
Km (Tm + m )
b = 078
. e1.9 + 1. 2 .
Ms = 2.0; 014
. m 55
.
Tm
2
3.4 me 0.28 0.0089
Table 14: PI tuning rules - FOLIPD model G m (s) =
Km e s
s(1 + sTm )
m
1
E (s ) Kc R( s) . 1 tuning rule.
U(s) = K c 1 +
Tis
Rule
Kc
Ti
0.049 Ku
2. 826 Tu
m
1.0 .
Tm
20% ; settling time
= 0.506 , c = 1640
0.066 Ku
2. 402 Tu
= 0.512 , c = 1600
0.099 Ku
1.962 Tu
= 0.522 , c = 1550
0.129 Ku
1.716 Tu
= 0.532 , c = 1500
0.159 Ku
1.506 Tu
= 0.544 , c = 1450
0.189 Ku
1.392 Tu
= 0.555 , c = 1400
0.218 K u
1.279 Tu
= 0.564 , c = 1350
0.250 Ku
1.216 Tu
= 0.573 , c = 1300
0.286 Ku
1.127 Tu
= 0.578 , c = 1250
0.330 Ku
1.114 Tu
= 0.579 , c = 1200
0.351K u
1.093Tu
= 0.577 , c = 1180
1
0 .2839
0 .1787 +
m
Km
+ 0. 001723
Tm
4. 296 + 3. 794 m + 0. 2591 m

Tm
Tm
= 0. 6551+ 0. 01877 m
Tm

[134b]
K m = 1 ; Tm = 1
Model: Method 4
Comment
Table 15: PI tuning rules - SOSPD model
K m e s
Km e s
(1 + Tm1s)(1 + Tm2 s)
m
Tm1 s2 + 2 m Tm1s + 1
2
or
1
- controller G c ( s) = Kc 1 +
. 11 tuning rules.
Ts
i
Kc
Ti
Comment
Tm1 + Tm 2 + 0.5 m
K m m ( 2 + 1)
Tm1 + Tm 2 + 05
. m
varies graphically with

m (Tm1 + Tm 2 ) -
Rule
Robust
Brambilla et al. [48]
m (Tm1 + Tm 2 )
Model: Method 1
0.1
0.2
0.5
Direct synthesis
Gain and phase margin - Tan
et al. [39] repeated pole
Model: Method 11
Regulator tuning
[59] page 158
Model: Open loop method
not specified
Model: Method 7

[17] page 48. Model:
method not specified.
m (Tm1 + Tm 2 )
3.0
1.8
1.0
Ti 1 + Tm
A m 1 + Ti
m (Tm1 + Tm 2 )
0.6
0.4
0.2
1.0
2.0
5.0
1
= 0.8,
< u
0.2
10.0
tan 2 tan 1 Tm m m
m
< 0.33 ;
Tm
= 0.5,
m
> 0.33
Tm

100Tm1
K m ( m
Minimum ISE McAvoy

and Johnson [83] deduced
m
from graph
1
1
Model: Method 1
1
Minimum ITAE Lopez et
al. [84] deduced from
graph
01
. m ( Tm1 + Tm 2 ) 10
T m1
+ Tm2 ) 50 + 551 e m + T m2
3 Tm1
+T
m 0.5 + 35
. 1 e ( m m2 )
Km
m Tm 1
m Tm 1
m Tm 1
0.5
4.0
10.0
0.8
5.7
13.6
1.82
12.5
25.0
4
4
0.5
4.0
4.3
27.1
3.45
6.67
7
7
0.5
4.0
7.8
51.2
3.85
5.88
Km
m
0.5
0.5
0.5
Tm
m Tm 1
0.1
3.0
1.0
0.2
10.0
0.3
0.77 Tm1
K m m
070
. Tm1
Km m
080
. Tm1
Km m
080
. Tm1
Km m
m Tm 1
m Tm 1
2.86
0.83
4.0
1
1
1
0.1
1.0
10.0
7.0
0.95
0.35
2.00
2.22
5.00
4
4
4
0.1
1.0
4.0
283
. ( m + Tm 2 )
2.65( m + Tm 2 )
2.29( m + Tm 2 )
167
. ( m + Tm2 )
m
Tm1
40.0 0.83
6.0 3.33
0.75 10.0
T
= 02
. , m2 = 01
.
Tm 1
m
T
= 02
. , m2 = 0.2
Tm1
Tm1
m
T
= 02
. , m2 = 05
.
Tm1
Tm 1
m
T
= 02
. , m2 = 10
.
Tm1
Tm 1
Rule
Kc

[18]
K c ( 31)
Ti
1
Comment
0<
Ti ( 31)
Tm 2
1 ; 01
. m 1
Tm1
Tm1
Model: Method 1
Kc (31) =
0.9077
0.063
m
m
1
Tm2
mTm2
6.4884 + 4.6198 m 3491
.
253143
.
+
0
.
8196
52132
.
2
Km
Tm 1
Tm1
Tm1
Tm1
Tm1
0.5961
0.7204
1.0049
1.005
m
Tm2
Tm2
Tm 2
1
T
7.2712
+
180448
.
+ 5.3263
+ 139108
.
+ 0.4937 m2
Km
m
Tm1
Tm1
Tm1
Tm1
1
+
Km
0.8529
0.5613
0.557
1.1818
Tm 2 m
Tm 2 m
m Tm2
m Tm 2
191783
.
+ 12.2494
+ 8.4355
17.6781
Tm1 Tm1
Tm1 Tm1
Tm1 Tm1
Tm1 Tm1
1
+
Km
Ti
( 31)
0.7241eT
m Tm 2
Tm 2
m1
2.2525e
Tm 1
+ 54959
.
e
Tm1 2
2
2
3
m
Tm2
m
m
Tm2
m Tm2
= Tm1 0.0064 + 3.9574

+ 4.4087
6.4789
.
15083
.
128702
+ 9.4348

2
Tm1
Tm1
Tm1
Tm1
Tm1
Tm1
T
+ Tm 1 17.0736 m2
Tm1
+ 15.9816 m
T
T
m1
m1
Tm2
T

. m 2 10.7619 m
3909
T
T
m1
m1
Tm1
3
2
2
T
T
+ Tm 1 10.684 m2 m 22.3194 m m 2 6.6602 m

Tm1 Tm1
Tm1
Tm1 Tm1
Tm2
T
+ 6.8122 m 2
Tm 1
Tm1
5
4
2
3
2
3
m
Tm2 m
m Tm 2
Tm2 m
+ Tm1 75146
.
.
.
.
+ 28724
+ 114666

+ 111207

Tm1 Tm1
Tm1
Tm1 Tm 1
Tm1 Tm1
4
5
6
5
6
m Tm 2
Tm 2
m
Tm2
Tm2 m
+ Tm 1 12174
.
.
.
4.3675
2.2236
0112
+ 10308

Tm1 Tm1
Tm1 Tm1
Tm1
Tm1
Tm1
4
2
3
3
2
4
T
m Tm2
m Tm2
m
+ Tm1 1. 9136 m m2 34994
.
15777
.
.

+ 11408
T
T
T
T
T
T
T
m1 m1
m1 m1
m1 m1
m1
Tm 2
Tm1
Rule
Kc

[18]
K c ( 32)
Ti
Comment
0.4 m 1 ;
Ti ( 32)
0.05
Model: Method 1
Kc
( 32)
1
=
Km
1.4439
0.1456
m
m
m
m
10.4183 209497
.
55175
.
m 265149
.
m
+ 42.7745
+ 105069
.
Tm1
Tm1
Tm1
Tm1
0.3157
0.0541

1
15.4103 m
+ 34.3236 m 3.7057 17.8860 m 4 .5359 54.0584 m 1.9593 + 22.4263 m m

Km
Tm1
Tm1
4. 7426

m
T
1
2.7497 m m
+ 50.2197 m 1.8288 m 171968
.
m 2.7227 + 10293
.
m m1
Km
Tm1
Tm1
Tm1
m
1
167667
.
e T + 14.5737e 7.3025 e
Km
Ti
( 32)
m
1
Tm1
m1
m
Tm1
2
3
m
m
m
m
2
= Tm1 11447
.
+ 45128
.
75.2486 m 110807
.
+ 345.3228 m + 191.9539
12.282 m

Tm1
Tm1
Tm1
Tm1
2
4
+ Tm1 359 .3345 m m 158.7611 m m 2 770.2897 m2 153633

. m
Tm1
Tm1
Tm1
3
2

m
+ Tm1 412 .5409 m m 414 .7786 m 2 m + 4850976
.
m 3 + 864.5195 m 4
Tm1
Tm1
Tm1
5
4
3
2

m
3
+ Tm 1 55.4366 m + 222 .2865 m m + 275166
. m 2 m + 2052493
.
Tm1
Tm1
Tm1
Tm1
6
5
m 4
m
m
5
6
+ Tm 1 479 .5627
.
m 6.547
.
m
m 4731346
432822
+ 99.8717 m
Tm1
Tm1
Tm1
4
3
2
m
m
m
2
3
4
5
+ Tm1 735666
.
56
.
4418
37
.
497
m + 160.7714 m m
Tm1
Tm1
Tm1
Tm1
Rule
Servo tuning
[18]
Model: Method 1
Kc (33) =
Kc
Ti
Comment

K c ( 33)
0<
Ti ( 33)
Tm 2
1 ; 01
. m 1
Tm1
Tm1
1.0169
3.5959

m
1
T
T
130454
.
9.0916 m + 2.6647 m2 + 9.162 m m2 2 + 0.3053 m
+ 11075
.
Km
Tm1
Tm 1
Tm1
Tm1
T
m
1
3.6843
0.8476
2.6083
2.9049

Tm2
T
T
1
T
2.2927 m
310306
.
13.0155 m2
+ 9.6899 m 2
0.6418 m 2
Km
m
Tm1
Tm1
Tm1
Tm1
0.2016
1.3293
0.801
1
Tm2 m
Tm 2 m
m Tm 2
189643
+
.
39.7340
+ 28155
.
Km
Tm1 Tm1
Tm1 Tm1
Tm 1 Tm1
Ti
( 33)
T
3.956
1
m Tm2
T
T
2.0067
+
+ 4.8259e + 2.1137e + 84511

.
eT
Km
Tm1 Tm1
2
2
3
m
Tm 2
m
m
Tm2
m Tm2
= Tm1 0.9771 0.2492
+ 0.8753
+ 3.4651
38516
.
+ 7.5106
7 .4538

2
Tm1
Tm1
Tm1
Tm1
Tm1
Tm1
2
2
3
4
Tm2 m
m Tm 2
Tm 2
m
+ Tm1 116768
.
.
.
10.9909
161461
+ 82567

Tm1 Tm1
Tm1 Tm1
Tm1
Tm1
m2
m1
m1
m2
2
m1
3
2
2
3
4
T
Tm 2
Tm2 m
m Tm 2
+ Tm 1 181011
.
.
.
+ 6.2208 m m2 + 219893
+ 158538

Tm1 Tm1
Tm 1 Tm1
Tm1 Tm1
Tm1
5
4
2
3
2
3

T
T
T
+ Tm1 4.7536 m + 14.5405 m 2 m 2.2691 m2 m 8.387 m m 2
Tm 1
Tm1 Tm1
Tm1 Tm1
Tm1 Tm1
4
5
6
5
6
m Tm2
Tm 2
m
Tm 2
Tm2 m
+ Tm 1 16.651
.
.
.
71990
+ 11496
4.728
+ 11395

Tm 1 Tm1
Tm1 Tm 1
Tm1
Tm1
Tm1
4
2
3
3
2
4
T
m Tm 2
m Tm2
+ Tm 1 0.6385 m m 2 + 10885
.
+
31615
.

+ 4.5398 m
T
T
T
T
T
T
T
m1 m1
m1 m1
m1 m1
m1
Tm 2
Tm1
Rule
Kc

[18]
K c ( 34 )
Ti
Comment
0.4 m 1 ;
Ti ( 34)
0.05
Model: Method 1
Kc
Ti
( 34)
( 34)
1
=
Km
m
1
Tm1
3
2
m
m
m
m
10.95 18845
.
3.4123 m + 4.5954 m
17002
.
.
m
21324

Tm1
Tm1
Tm1
Tm 1
0.421
0.1984
1.8033

m
1
14.4149 m 2 m 0.7683 m 3 + 7.5142 m
+ 3.7291 m
+ 53444
.
Km
Tm1
Tm1
Tm1
Tm1
0.6753
0.1642

1
0.0819 m 19 .5419 3603
. m 1.0749 + 71163
.
m 1.1006 + 3206
. m m
7.8480 m m
Km
Tm1
Tm1

1
113222
.
m1.9948 m + 2.4239e T
Km
Tm1
m1
+ 34137
.
e + 10251
.
e
m
m m
Tm1
05593
.
m
Tm1
2
3
m
m
m
m
2
= Tm1 2.4866 233234

.
+ 53662
.
m + 656053
.
24.1648 m 83.6796
+ 29.0062 m

Tm1
Tm1
Tm1
Tm1
2
4

m

+ Tm1 1359699
.
m m + 431477
.
m 2 + 519749
.
m 3 + 86.0228 m
Tm1
Tm 1
Tm1
3
2

m
+ Tm 1 704553
.
m m + 1534877
.
m 2 m 1250112
.
m 3 685893
.
m4
Tm1
Tm1
Tm1
5
4
3
2

+ Tm 1 62.7517 m + 27.6178 m m 152 .7422 m 2 m + 20.8705 m m 3
Tm1
Tm1
Tm1
Tm 1
6
5
m 4
m
m
5
+ Tm1 54.0012
.
.
m
.
m 6
m + 58.7376 m + 131193
+ 202645
232064
Tm1
Tm1
Tm1
4
3
2
m
m
m
m
2
3
4
+ Tm 1 616742
.
+
136
.
2439
954092
.
.
m 5
m + 204168
Tm1
Tm1
Tm1
Tm1
Kc
Rule
Ti
Comment
Ultimate cycle
Decay ratio = 0.15 - < 2.4 ,
Regulator - nearly minimum
[60]
Kc
( 7)
Ti
( 7)
0.2
m
Tm1
2.0 , 0.6 m 4.2
Decay ratio = 0.15 -
Kc( 8)
Model: Method 3
Ti (8)
2.4 < 3 , 0.2
m
Tm1
2.0 ,
0.6 m 4.2
Kc ( 9 ) 5
Ti ( 9)
3 < 20 , 0.2
m
Tm1
2.0 ,
0.6 m 4.2
6Tm1 + 4 m Tm1 m + K H Km m
9
, KH =
2
2
2 Tm1 m H
2 m K m
2
H =
Tm12
Kc
( 7)
Kc
(8 )
Kc
( 9)
1 + KH K m
2T
K K 2
+ m1 m m + H m m
3
6
0.674 1 0.447 H m + 0.0607 ( H m )
= 1
K H K m (1 + K H K m )
0.778 1 0.467 H m + 0.0609( H m )
= 1
K H K m (1 + K H K m )
2
m 2
m Tm1 m
2
,
m1
18
18
] K
, Ti ( 7) =
, Ti (8 ) =
] K
K c (1 + K H K m )
0.0607 H K m 1 + 1.05 H m 0.233 H m
K c (1 + K H K m )
0.0309 H K m 1 + 2.84 H m 0.532 H m

2
131
. ( 0519
. )
1 103
. + 0514
.
2
K c (1 + KH K m )
( 9)
= 1
K H , Ti =
KH Km 1 + KH K m
0.0603 1 + 0.929 ln[ H m ] 1 + 2.01 12
. 2
)(
)
)
)
Kc
Rule
Kc
Ti
( 10 ) 6
Ti
Comment
Decay ratio = 0.15 - > 20 ,
( 10)
0.2

IAE, ISE, ITAE - Hwang [60]
m
Tm1
2.0 , 0.6 m 4.2
Decay ratio = 0.2 - < 2.4 ,
Kc
Model: Method 3
( 11)
Ti
( 11)
Ti
( 12)
0.2
m
Tm1
2.0 , 0.6 m 4.2
Decay ratio = 0.2 -
Kc
( 12 )
2.4 < 3 , 0.2
2.0 ,
Tm1
0.6 m 4.2
Decay ratio = 0.2 -
Kc(13 )
Ti (13)
3 < 20 , 0.2
2.0 ,
Tm1
0.6 m 4.2
Decay ratio = 0.2 - > 20 ,
Kc(14 )
Ti (14)
Kc(15)
Ti (15)
(16 )
( 16)
0.2
m
Tm1
2.0 , 0.6 m 4.2
Decay ratio = 0.25 - < 2.4 ,

0.2
m
Tm1
2.0 , 0.6 m 4.2
Kc
Ti
2.4 < 3 , 0.2
2.0 ,
Tm1
0.6 m 4.2
Kc
( 10)
Kc
( 11)
Kc
( 12 )
114
. 1 0.482 H m + 0.068( H m )
= 1
K H K m (1 + K H K m )
] K
0.622 1 0.435 H m + 0.052( H m )
= 1
K H K m (1 + K H K m )
] K
0.724 1 0.469 H m + 0.0609 ( H m )
= 1
K H K m (1 + K H K m )
, Ti (10) =
, Ti (11) =
] K
K c (1 + K H K m )
2
0.0697 H K m 1 + 0.752 H m 0.145 H m
0.0694 H K m 1 + 2.1 H m 0.367 H m
, Ti (12 ) =
K c (1 + K H K m )
2
K c (1 + K H K m )
0.0405 H K m 1 + 1.93 H m 0.363 H m

2
126
. (0.506)
1 107
. + 0.616 2
K c (1 + K H K m )
Kc (13) = 1
K H , Ti (13) =
K H Km 1 + KH K m
0.0661(1 + 0.824 ln[ H m ])(1 + 171
. 117
. 2 )
Kc
( 14 )
Kc
( 15)
Kc
( 16)
109
. 1 0.497 H m + 0.0724( H m )
= 1
K H K m (1 + K H K m )
] K
0.584 1 0.439 H m + 0.0514 ( H m )
= 1
K H K m (1 + K H K m )
0.675 1 0.472 H m + 0.061( H m )
= 1
K H K m (1 + K H K m )
] K
] K
, Ti (14 ) =
0.054 H K m 1 + 2.54 H m 0.457 H m
, Ti (15) =
, Ti (16) =
K c (1 + K H K m )
2
K c (1 + K H K m )
0.0714 H K m 1 + 0.685 H m 0.131 H 2 m 2
K c (1 + K H K m )
0.0484 H K m 1 + 1.43 H m 0.273 H m

2
Rule
Kc
Ti
Regulator - nearly minimum

IAE, ISE, ITAE - Hwang [60]
Kc(17 ) 7
Ti (17)
Comment
Decay ratio = 0.25 m
3 < 20 , 0.2
2.0 ,
Tm1
0.6 m 4.2
Model: Method 3
Decay ratio = 0.25 - > 20 ,
Kc (18 )
Ti (18)
0.2
m
Tm1
2.0 , 0.6 m 4.2
Decay ratio = 0.1 - < 2.4 ,

Servo - nearly minimum IAE,
ISE, ITAE - Hwang [60]
Kc
( 19 )
Ti
0.2
( 19)
m
Tm1
2.0 , 0.6 m 4.2
m 0.776 + 0.0568
Model: Method 3

m
+ 018
. m
Tm1
T m1
Decay ratio = 0.1 - 2.4 < 3 ,
Kc
( 20)
Ti
0.2
( 20)
m
Tm1
2.0 , 0.6 m 4.2
m 0.776 + 0.0568

m
+ 018
. m
Tm1
T m1
Decay ratio = 0.1 - 3 < 20 ,
Kc
( 21)
Ti
0.2
( 21)
m
Tm1
2.0 , 0.6 m 4.2
m 0.776 + 0.0568

m
+ 018
. m
Tm1
T m1
Decay ratio = 0.1 - > 20 ,
Kc
( 22 )
Ti
0.2
( 22 )
m
Tm1
2.0 , 0.6 m 4.2
m 0.776 + 0.0568
Kc
( 17)
Kc
( 18)
Kc
( 19 )
Kc
( 20)

2
12
. ( 0495
. )
1 11
. + 0.698
Kc (1 + KH K m )
K , T (17 ) =
= 1
i
H
0.0702 1 + 0.734 ln[ H m ] 1 + 148

. 11
. 2
KH Km (1 + K H Km )

m
+ 018
. m
Tm1
T m1
1.03 1 0.51 H m + 0.0759( H m )
= 1
K H K m (1 + K H K m )
] K
0.822 1 0.549 H m + 0.112( H m )
= 1
K H K m (1 + K H K m )
] K
0.786 1 0.441 H m + 0.0569( H m )
= 1
K H K m (1 + K H K m )
)(
, Ti (18) =
] K
0.0386 H K m 1 + 3.26 H m 0.6 H m
, Ti (19 ) =
K c (1 + K H K m )
2
, Ti ( 20) =
K c (1 + K H K m )
0.0142 H K m 1 + 6.96 H m 177

. H m
2
K c (1 + K H K m )
0.0172 H K m 1 + 4.62 H m 0.823 H m

2

128
. ( 0542
. )
1 0.986 + 0558
.
2
K c (1 + K H K m )
( 21)
Kc ( 21) = 1
KH , Ti = 0.0476 1 + 0.996 ln 1 + 213
K
K
1
+
K
K
. 2 )
(
[ H m ])( . 113
H
m
H m
H
Kc
( 22)
114
. 1 0.466 H m + 0.0647 ( H m )
= 1
K H K m (1 + K H K m )
] K
, Ti (22) =
K c (1 + K H K m )
0.0609 H K m 1 + 1.97 H m 0.323 H 2 m2
Kc
Rule
Ti
Comment
Decay ratio = 0.1 - < 2.4 ,

K c ( 23)
0.2
Ti ( 23)
m
Tm1
2.0 , 0.6 m 4.2
> 0889
.
+ 0.496
Model: Method 3

m
+ 0.26 m
Tm1
Tm1
Decay ratio = 0.1 - 2.4 < 3 ,

K c ( 24 )
0.2
Ti ( 24)
m
Tm1
2.0 , 0.6 m 4.2
> 0889
.
+ 0.496

m
+ 0.26 m
Tm1
Tm1
Decay ratio = 0.1 - 3 < 20 ,

Kc
( 25)
Ti
0.2
( 25)
m
Tm1
2.0 , 0.6 m 4.2
> 0889
.
+ 0.496

m
+ 0.26 m
Tm1
Tm1
Decay ratio = 0.1 - > 20 ,

K c (26)
0.2
Ti (26)
m
Tm1
2.0 , 0.6 m 4.2
> 0889
.
+ 0.496

m
+ 0.26 m
Tm1
Tm1
Decay ratio = 0.1 - < 2.4 ,

K c ( 27 )
0.2
Ti ( 27)
m
Tm1
2.0 , 0.6 m 4.2

m
+ 018
. m
Tm1
T m1
m > 0.776 + 0.0568

m 0.889 + 0.496
Kc
( 23)
Kc
( 24)
Kc
( 25)
2
0.794 1 0541
. H m + 0.126( H m )
= 1
K H K m (1 + KH K m )
0.738 1 0.415 + 0.0575 2

( H m)
H m
= 1
KH K m (1 + KH K m )
Kc
, Ti (23) =
] K
, Ti (24) =
K c (1 + K H K m )
0.0078 H K m 1 + 8.38 H m 197

. H2 m 2
K c (1 + K H K m )
0.0124 H K m 1 + 4.05 H m 0.63 H 2 m 2

2
115
. (0.564) H m 1 0.959 + 0.773
K c (1 + K H K m )
K , T (25) =
= 1
H i
K
K
(
1
+
K
K
)
0
.
0355
1
+
0
.
947
ln[ H m ] 1 + 19
. 107
. 2
H m
H m
107
. 1 0.466 H m + 0.0667( H m )
Kc ( 26) = 1
K H Km (1 + K H Km )
( 27)
] K
0.789 1 0.527 + 0.11 2

( H m)
H m
= 1
K H K m (1 + K H K m )
] K
] K
)(
, Ti (26) =
, Ti (27) =

m
+ 0.26 m
Tm1
Tm1
K c (1 + K H K m )
0.0328 H K m 1 + 2.21 H m 0338

. H2 m 2
K c (1 + K H K m )
0.009 H K m 1 + 9.7 H m 2.4 H 2 m 2
Kc
Rule
Ti
Comment
Decay ratio = 0.1 - 2.4 < 3 ,

K c (28)
0.2
Ti (28)
m
Tm1
2.0 , 0.6 m 4.2

m
+ 018
. m
Tm1
T m1
m > 0.776 + 0.0568
Model: Method 3
m 0.889 + 0.496

m
+ 0.26 m
Tm1
Tm1
Decay ratio = 0.1 - 3 < 20 ,

K c ( 29 )
0.2
Ti ( 29)
m
Tm1
2.0 , 0.6 m 4.2

m
+ 018
. m
Tm1
T m1
m > 0.776 + 0.0568

m 0.889 + 0.496

m
+ 0.26 m
Tm1
Tm1
Decay ratio = 0.1 - > 20 ,

K c ( 30)
0.2
Ti (30)
m
Tm1
2.0 , 0.6 m 4.2

m
+ 018
. m
Tm1
T m1
m > 0.776 + 0.0568

m 0.889 + 0.496
Regulator - minimum IAE Shinskey [17] page 121.

Model: method not
specified
Kc
( 28)
048
. Ku
m
T
= 02
. , m2 = 0.2
Tm1
Tm1
083
. Tu
2
0.76 1 0426
. H m + 0.0551( H m )
= 1
K H Km ( 1 + K H Km )
] K
, Ti (28) =

m
+ 0.26 m
Tm1
Tm1
K c (1 + K H K m )
0.0153 H K m 1 + 4.37 H m 0.743 H 2 m 2
Kc
( 29)

2
122
. (0.55) H m 1 0.978 + 0.659
K c (1 + K H K m )
K , T (29) =
= 1
H i
K
K
(
1
+
K
K
)
0.0421 1 + 0.969 ln[ H m ] 1 + 2.02 111
. 2
H m
H m
Kc
( 30)
2
111
. 1 0.467 H m + 0.0657( H m )
= 1
KH K m ( 1 + K H Km )
] K
)(
, Ti (30) =
K c (1 + K H K m )
0.0477 H K m 1 + 2.07 H m 0.333 H2 m 2
Table 16: PI tuning rules - SOSPD model -
K m e s
Tm1 s2 + 2 m Tm1s + 1
2
- Two degree of freedom controller: U(s) = K c 1 + 1 E (s ) Kc R( s) . 3 tuning rules.
Kc
Rule
Ti
Tis
1
0.5613
0
.
3717
+
m
K m
+ 0 .0003414
Tm
Ti
1
Km
m
+ 0. 3087 m 0. 1201 m
Tm
Tm
Tm
0
.
05627
0
.
1000
+
m
2
[
+ 0.06041]
Ti
m
1.0 ; m = 1
Tm
20% ; settling time
( 30a ) 10
= 0. 6438 0. 5056
= 0. 6178 0. 4439 m 7. 575 m + 9. 317 m

Tm
T
m
Tm

[134a]
Comment
( 30b )
3. 182 m
m
1.0 ; m = 0.5
Tm
20% ; settling time
= 0. 4002 ,
c = 139.65 0
K m = 1 ; Tm = 1 ; m = 1
0.1713 Ku
1.0059 Tu
0.147 K u
1.150 Tu
= 0. 411 , c = 1460
0.170 K u
1.013Tu
= 0. 401 , c = 1400
0.195 K u
0.880 Tu
= 0.386 , c = 1330
0.210 K u
0.720 Tu
= 0.342 , c = 1250
0.234 K u
0.672 Tu
= 0.345 , c = 1150
0.249 K u
0.610 Tu
= 0.323 , c = 1050
0.262 K u
0.568 Tu
= 0.308 , c = 940
0.274 K u
0.545 Tu
= 0. 291 , c = 840
0.280 K u
0.512 Tu
= 0. 281 , c = 730
0.291K u
0.503 Tu
= 0.270 , c = 630
0.297 K u
0.483 Tu
= 0.260 , c = 520
0.303 K u
0.462 Tu
= 0.246 , c = 410
0.307 K u
0.431Tu
= 0.229 , c = 300
Model: Method 15
[134b]
K m = 1 ; Tm = 1 ; m = 1
Model: Method 15
2
3
m
m
m
= Tm 2.069 0.3692
+ 1.081 0.5524
Tm
Tm
Tm
2
3
4

= Tm 4. 340 16. 39 m + 30. 04 m 25 .85 m + 8.567 m
Tm
Tm
Tm
Tm
( 30a )
10
Ti
Ti
( 30b )
Rule
Kc
Ti
Comment

[134b] - continued
0.317 K u
0.386 Tu
= 0. 171 , c = 190
0.324 K u
0.302 Tu
= 0.004 , c = 100
0.320 K u
0.223 Tu
= 0.204 , c = 60
Table 17: PI tuning rules - SOSIPD model (repeated pole) -
K m e s
s (1 + Tm1s) 2
- Two degree of freedom controller: U(s) = K c 1 + 1 E (s ) Kc R( s) . 1 tuning rule.
Rule
Kc
Ti
K m = 1 ; Tm = 1
Model: Method 1
Comment
m
Tm
m
1.0 ;
Tm
20% ; settling time
0.049 Ku
2. 826 Tu
= 0.506 , c = 1640
0.066 Ku
2. 402 Tu
= 0.512 , c = 1600
0.099 Ku
1.962 Tu
= 0.522 , c = 1550
0.129 Ku
1.716 Tu
= 0.532 , c = 1500
0.159 Ku
1.506 Tu
= 0.544 , c = 1450
0.189 Ku
1.392 Tu
= 0.555 , c = 1400
0.218 K u
1.279 Tu
= 0.564 , c = 1350
0.250 Ku
1.216 Tu
= 0.573 , c = 1300
0.286 Ku
1.127 Tu
= 0.578 , c = 1250
0.330 Ku
1.114 Tu
= 0.579 , c = 1200
0.351K u
1.093Tu
= 0.577 , c = 1180
1
0 .3840
0
.
07368
+
m
K m
+ 0.7640
Tm
8.549 + 4.029 m
Tm
= 0. 6691 + 0.006606

[134b]
Tis
Km e s
- controller G c ( s) = Kc 1 + 1 . 1
Ts
(1 + sTm1 )(1 + sTm2 )(1 + sTm 3 )
i
tuning rule
m
Table 19: PI tuning rules - TOLPD model
Rule
Hougen [85]
Kc
0.7 Tm1
Km m
Ti
15
. m
0.08
Tm1( Tm 2 + Tm3 )
15
. m
0.08
Tm1( Tm 2 + Tm3 )
m
0.04 ; Tm1 Tm 2 Tm3
Tm1
Model: Method 1
0. 333
1 Tm 1
T + Tm2 + Tm 3
0.7
+ 0.8 m1
0. 333
2K m m
( Tm 1Tm 2 Tm 3 )
Comment
m
> 0.04 ; Tm1 Tm 2 Tm3
Tm1
0.333
Table 20: PI tuning rules - TOLPD model (repeated pole) -
K m e s
(1 + Tm1s ) 3
- Two degree of freedom controller: U(s) = K c 1 + 1 E (s ) Kc R( s) . 1 tuning rule.
Rule
Kc
Ti
Ti
( 30c )
Comment
1
0.7399
0
.
2713
+
m
K m
+ 0.5009
Tm
Ti
( 30 c ) 1
= 0. 4908 0. 2648 m + 0. 05159 m

Tm
Tm
Tis
= Tm 2.759 0. 003899 m + 0.1354 m
Tm
Tm
m
1.0 ;
Tm
20% ; settling time
K me s
1 sTm
Table 21: PI tuning rules - unstable FOLPD model
Kc
Rule
Ts
i
Ti
Comment
Direct synthesis
1
De Paor and OMalley [86]
Kc
( 35)
Model: Method 1
Venkatashankar and
Chidambaram [87]
Model: Method 1
Chidambaram [88]
Model: Method 1
1 Tm m
Tm
tan( 0.5m )
Tm m
Kc( 36)
25( Tm m )
1
T
1 + 0.26 m
Km
m
25Tm 27 m
p Tm
Ho and Xu [90]
A m Km
Model: Method 1
m
<1
Tm
gain margin = 2;
m = tan 1
1 Tm m
T m m 1 Tm m
Tm m
m
< 0.67
Tm
m
< 06
.
Tm
p =
1
1.57 p p m
2
1
Tm
A m m + 157
. A m (A m 1)
(A
2
m
1 m
m
< 0.62
Tm
Robust
Rotstein and Lewin [89]
Tm
+ 2
Tm
2
Km
+ 2
Tm
Km uncertainty = 50%
m Tm = 0.2
= [0.6Tm ,1.9Tm ]
m Tm = 0.2
= [ 0.5Tm , 45
. Tm ]
m Tm = 0.4
= [15
. Tm ,4.5Tm ]
m Tm = 0.6
= [ 39
. Tm ,41
. Tm ]
Model: Method 1
obtained graphically
sample values below
Ultimate cycle
Luyben [91]
031
. Ku

modulus = 2 dB; closed
loop time constant =
316
. m
2.2Tu
Kc
Kc
( 35)
( 36)
Tm
1 Tm m
sin
cos (1 Tm m )Tm m +
K m
Tm m
1
=
Km
2
0.98 1 + 0.04Tm
2
( Tm m )
(1 Tm m )Tm m
2 Tm2 (Tm m )
1+

(Tm m ) 2 2m :
m
25 ( T )
< 025
.
= 1373
. ,
m
m
m
625
2
Tm
1+2
T
(
)
m
m
2
m
= 0.953, 0.25
m
< 0.67
Tm
K m e s
- controller G c ( s) = Kc 1 + 1 . 3
Ts
(1 sTm1 )(1 + sTm 2 )
i
tuning rules.
m
Table 22: PI tuning rules - unstable SOSPD model
Rule
Kc
Ti
Comment
Ultimate cycle
McMillan [58]
Kc( 37) 3
Ti (37 )

from Ku , Tu
Regulator tuning

Pomerleau [82] deduced
from graph
Model: Method 1
Output step load
disturbance
(considered as 2 tuning
rules)
Input step load disturbance
bTm1 1 +
4Tm1 ( m + Tm2 )
aTm2 2
4( m + Tm 2 )
aTm1 4( m + Tm2 )
K m aTm1 4[ m + Tm2 ]
m Tm
m Tm
m Tm
0.05
0.10
0.15
0.20
0.9479
1.0799
1.2013
1.3485
2.3546
2.4111
2.4646
2.5318
0.25
0.30
0.35
0.40
1.4905
1.6163
1.7650
1.9139
2.5992
2.6612
2.7368
2.8161
0.45
0.50
2.0658
2.2080
2.9004
2.9826
0.05
0.10
0.15
0.20
1.1075
1.2013
1.3132
1.4384
2.4230
2.4646
2.5154
2.5742
0.25
0.30
0.35
0.40
1.5698
1.6943
1.8161
1.9658
2.6381
2.7007
2.7637
2.8445
0.45
0.50
2.1022
2.2379
2.9210
3.0003
Kc ( 37) =
1477
.
Tm1Tm 2
K m m2
Ti A max
Km
Tm1 Tm2 max + Tm1 + Tm 2 max + max

2
Kc ( 38) =
0.65
(T + T )T T
m1
m2
m1 m 2
( 37)
,
T
=
332
.
1
+
i
m
0.65
( Tm1 + Tm 2 )Tm1Tm 2
(Tm1 Tm 2 )(Tm1 m ) m

1+
(Tm1 Tm2 )(Tm1 m ) m
2
1 + Ti max
2

1
Table 23: PI tuning rules delay model e s m - controller G c ( s) = Kc 1 +
. 2 tuning rules
Ts
i
Rule
Direct synthesis
Hansen [91a]
Model: Method not
specified
Regulator tuning
Shinskey [57] minimum
IAE regulator - page 67.
Model: method not
specified
Kc
Ti
0.2
0.3m
0.4
Km
0.5 m
Comment
K m e s
1
- ideal controller G c ( s) = Kc 1 +
+ Td s . 56.
1 + sTm
Ti s
Table 25: PID tuning rules - FOLPD model
Kc
Rule
Process reaction
Ziegler and Nichols
[8]
Model: Method 1
Astrom and
Hagglund [3] page
139
Model: Method 6
Parr [64] page 194
Model: Method 1
Chien et al. [10]
regulator
Model: Method 1
Chien et al. [10] servo

Model: Method 1
Three constraints
method - Murrill [13]page 356
Ti
Td
Comment
2 m
05
. m
Quarter decay ratio
0.94Tm
K m m
2 m
05
. m
125
. Tm
K m m
2.5 m
0. 4 m
095
. Tm
Km m
2.38 m
0.42 m
0% overshoot;
011
. < m <1
Tm
12
. Tm
K m m
2 m
0.42 m
20% overshoot;
011
. < m <1
Tm
0.6Tm
K m m
Tm
0.5 m
0% overshoot;
011
. < m <1
Tm
095
. Tm
Km m
1.36Tm
0.47 m
20% overshoot;
011
. < m <1
Tm
12
. Tm 2Tm
,
]
K m m K m m
Tm
1370
.

Km m
0.950
Tm m

1351
.
Tm
0.738

0.365Tm m
Tm
0.950
Model: Method 3
Quarter decay ratio;

minimum integral error
(servo mode);
Kc K m Td
= 05
. ;
Tm
01
.
Cohen and Coon [11]

Model: Method 1
Astrom and
Hagglund [93]pages 120-126

2.5 m + 0.46 m
1
Tm
Tm
.
+ 025
. T Tm
135
Km
m
m
m
1 + 0.61
Tm
3
Km
Model: Method 19
Sain and Ozgen [94] 1
T
.
0. 6939 m + 01814
Model: Method 5 K m
m
Regulator tuning
Minimum IAE
Murrill [13] pages
358-363
Model: Method 3
0.37 m
1 + 0.2 m
Tm
m
1
Tm
Quarter decay ratio
T90% = 90% closed

loop step response
time. Leeds and
Northrup Electromax V.
T90%
05
. m
0.8647Tm + 0.226 m
Tm
+ 0.8647
m
0.0565Tm
T
0.8647 m + 0.226
m
Tm
1435
.

Km m
0.921
Tm m

0878
.
Tm
0.749

0.482 Tm m
Tm
1.137
01
. <
m
1
Tm
Rule
Modified minimum
IAE - Cheng and
Hung [95]
Model: Method 7
Minimum ISE Murrill [13] pages
358-363
Model: Method 3
Atherton [20]
Kc
3 Tm

K m m
Ti
0 .921
Td
Tm m

0878
.
Tm
Comment
0.749

0.482 Tm m
Tm
1.137
Tm
1495
.

Km m
0.945
Tm m

1101
.
Tm
0.771

0.56Tm m
Tm
1. 006
Tm
1473
.

Km m
0.970
Tm m

1115
.
Tm
0.753

0.55Tm m
Tm
0.948
Tm
1524
.

Km m
0.735
Tm m

1130
.
Tm
0.641

0.552Tm m
Tm
0.851
Tm
1357
.

Km m
0 .947
Tm m

0842
.
Tm
0.738

0.381Tm m
Tm
0.995
Tm
1468
.

Km m
0.970
Tm m

0942
.
Tm
0.725

0.443Tm m
Tm
0.939
Tm
1515
.

Km m
0.730
Tm m

0957
.
Tm
0.598

0.444 Tm m
Tm
0 .847
Model: Method 6
Minimum ISTES Zhuang and
Atherton [20]
Tm
1531
.

Km m
0.960
Tm m

0971
.
Tm
0.746

0.413Tm m
Tm
0.933
Tm
1592
.

Km m
0.705
Tm m

0957
.
Tm
0.597

0.414 Tm m
Tm
0.850
Model: Method 6
Model: Method 6
Minimum ITAE Murrill [13] pages
358-363
Model: Method 3
Minimum ISTSE Zhuang and
Atherton [20]
Minimum error - step

load change - Gerry
[96]
Model: Method 6
Servo tuning
Minimum IAE Rovira et al. [21]
Model: Method 3
03
.
Km
Tm
1086
.

Km m
0.5m
0.869
0.6032
0.7645 +
(Tm + 05
. m)
m Tm
Tm + 0.5 m
Model: Method 6
Minimum ISE - Wang
et al. [97]
0.7524
0.9155 +
(Tm + 0 .5 m )
m Tm
Tm + 0.5 m
K m ( Tm + m )
K m (Tm + m )
Model: Method 6
Atherton [20]
Model: Method 6
Tm
1048
.

Km m
0.897
Tm
1154
.

K m m
0.567
0.5Tm m
Tm + 05
. m
0.5Tm m
Tm + 05
. m
Tm

0.489 Tm m
Tm
0.888
Tm

0.490 Tm m
Tm
0.708
1195
.
0.368 m
Tm
1.047 0.220 m
Tm
m
1
Tm
01
.
m
10
.
Tm
11
.
m
2.0
Tm
01
. <
m
1
Tm
01
.
m
10
.
Tm
11
.
m
2.0
Tm
01
.
m
10
.
Tm
11
.
m
2.0
Tm
m
>5
Tm
Tm

0.914
Tm
m
0.348 Tm
0.740 0.13 m
Tm
Tm
Minimum IAE Wang et al. [97]
01
. <
01
. <
m
1
Tm
0.05 <
m
<6
Tm
0.05 <
m
<6
Tm
01
.
m
10
.
Tm
11
.
m
2.0
Tm
Rule
Kc
Ti
Minimum ITAE Rovira et al. [21]

Model: Method 3
Tm
0965
.

K m m
Modified minimum
ITAE - Cheng and
Hung [95]
Model: Method 7
Minimum ITAE
Wang et al. [97]
12
. Tm

Km m
0.85
0.855
Td
Tm

0.308 Tm m
Tm
Tm

0.308 Tm m
Tm
0.929
0.796 0.1465 m
Tm
0.796 0.147 m
Tm
05307
.
0.7303 +
(Tm + 0.5 m )
m Tm
K m (Tm + m )
Tm + 0.5 m
0.5Tm m
Tm + 05
. m
Model: Method 6
Minimum ISTSE
Zhuang and
Atherton [20]
Model: Method 6
Minimum ISTES
Zhuang and
Atherton [20]
Model: Method 6
Tm
1042
.

Km m
0 .897
Tm
1142
.

Km m
0.579
Tm
0968
.

Km m
0.904
Tm
1061
.

Km m
0 .583
Comment
0.929
Tm

0.385Tm m
Tm
0.906
Tm

0.384 Tm m
Tm
0 .839
Tm

0.316 Tm m
Tm
0.892
Tm

0.315Tm m
Tm
0.832
0.987 0.238 m
Tm
0.919 0.172
m
Tm
0.977 0.253 m
Tm
0.892 0.165 m
Tm
Ultimate cycle
Regulator minimum
IAE Pessen [63]
0.7K u
0.4Tu
Model: Method 6
Servo minimum
0.051( 3. 302 K m Ku + 1) Tu
ISTSE Zhuang and
0.509Ku
Atherton [20]
Model: Method not
relevant
Servo minimum
ISTSE Pi-Mira et
0.604 K u
0.04 (4. 972 K m K u + 1)T
al. [97a]
Model: Method 27
Regulator - minimum
ISTSE - Zhuang and 4.434 K m K u 0.966
1751
. K m K u 0.612
Ku
Tu
Atherton [20]
512
. K m K u + 1734
.
3776
. K m K u + 1388
.
Model: Method not
relevant
0149
. Tu
0125
. Tu
01
.
m
1
Tm
Damping factor of
closed loop system =
0.707.
0.05 <
m
<6
Tm
01
.
m
10
.
Tm
11
.
m
2.0
Tm
01
.
m
10
.
Tm
11
.
m
2.0
Tm
01
.
m
10
.
Tm
01
.
m
2.0
Tm
01
.
m
2.0
Tm
0.130 Tu
0144
. Tu
Kc
Rule
Ti
Td
Ti (38 )
0471
. Ku
K m u
Comment
Regulator - nearly
minimum IAE, ISE,
ITAE - Hwang [60]
Kc( 38)
Model: Method 8
Decay ratio = 0.15 0.1
m
2.0 ; < 2.4
Tm
Model: Method 8
Kc
( 39)
Ti
( 39 )
0471
. Ku
K m u
0.1
m
2.0 ;
Tm
2.4 < 3
Kc( 40)
Ti (40 )
0471
. Ku
K m u
0.1
m
2.0 ;
Tm
3 < 20
Kc( 41)
Ti (41)
0471
. Ku
K m u
( 42)
( 42 )
0471
. Ku
K m u
Ti (43)
0471
. Ku
K m u
0.1
m
2.0 ; 20
Tm
Decay ratio = 0.2 -
Kc
Ti
0.1
m
2.0 ; < 2.4
Tm
Decay ratio = 0.2 -
Kc( 43)
0.1
m
2.0 ;
Tm
2.4 < 3
Kc
Rule
KH =
H =
Kc
( 38)
Kc
( 39)
Ti
2
9
. Km Kc Ku m
m m Tm + 1884
+
2
18
9u
2 K m m 18
Td
m4 49 m 2Tm 2 7Tm m 3 0.471Ku K c

+
+
324
324
162
u
1+ KH Km
m Tm K H K m m
0.942 K c K u m
+
3
6
3 u
2
2
0.674 1 0.447 H m + 0.0607 H m
= KH
K m (1 + K H K m )
2
2
0.778 1 0.467 H m + 0.0609 H m

= KH
K m (1 + KH K m )
10 m3 16 Tm m2 1775
. Kc 2 Ku 2 m2
+
81
81 u
81
2Tm u + K H K m m u 1884
. Ku Kc
0471
. K c K u H m
,=
Comment
] , T
] , T
( 38)
( 39)
Kc
( 38)
(1 + KH K m )
H Km 0.0607 1 + 105
. H m 0233
. H 2 m2
Kc
( 39)
(1 + K H Km )
H K m 0.0309 1 + 2.84 H m 0532

. H2 m 2

( 40)
131
. ( 0519
. ) H m 1 103
. + 0.514 2
K c (1 + K H K m )
Kc ( 40) = K H
, Ti (40) =
K m (1 + K H Km )
H K m 00603
.
1 + 0.929 ln[ H m ] 1 + 2.01 12
. 2
114
. 1 0.482 H m + 0.068 H 2 m 2
Kc ( 41) = KH
K m (1 + K H Km )
] , T
0622
.
1 0.435 H m + 0.052 H m
Kc ( 42) = K H
K m (1 + K H K m )
)(
( 41)
] , T
0.724 1 0469
. H m + 00609
.
H 2 m2
Kc ( 43) = KH
K m (1 + K H K m )
( 42)
] , T
Kc
( 41)
(1 + KH K m )
H K m 0.0694 1 + 21
. H m 0.367 H 2 m2
=
( 43)
Kc
( 42)
(1 + KH K m )
H K m 00697
.
1 + 0752
. H m 0.145 H 2 m 2
=
Kc
( 43)
(1 + K H K m )
H K m 0.0405 1 + 193
. H m 0.363 H 2 m 2
)
)
)
)
Regulator nearly
minimum IAE, ISE,
ITAE - Hwang [60]
(continued)
Kc
Model: Method 8
Decay ratio = 0.2 -
( 44)
Ti
( 44 )
Kc( 45)
Ti (45)
Kc
( 46)
Ti
( 46 )
Kc
( 47)
Ti
( 47 )
0471
. Ku
K m u
0.1
m
2.0 ;
Tm
3 < 20
Decay ratio = 0.2 -
0471
. Ku
K m u
m
2.0 ; 20
Tm
0.1
0471
. Ku
K m u
0.1
m
2.0 ; < 2.4
Tm
0471
. Ku
K m u
0.1
m
2.0 ;
Tm
2.4 < 3
Kc( 48)
Servo - nearly min.

IAE, ISE, ITAE Hwang [60]
Kc
( 49)
Kc
( 5 0)
Ti (48 )
0471
. Ku
K m u
Ti
( 49 )
0471
. Ku
K m u
Ti
( 50)
0471
. Ku
K m u
0.1
m
2.0 ;
Tm
3 < 20
Decay ratio = 0.25 m

2.0 ; 20
Tm
0.1
Decay ratio = 0.1 01

.
m
2.0 ; < 2.4
Tm
Model: Method 8

( 44)
126
. ( 0.506) H m 1 107
. + 0616
.
2
K c (1 + K H K m )
Kc ( 44) = K H
, Ti (44) =
K m (1 + K H K m )
H K m 00661
.
1 + 0.824 ln[ H m ] 1 + 171
. 117
. 2
Kc
( 45)
109
. 1 0.497 H m + 0.0724 H2 m 2
= KH
K m (1 + K H Km )
] , T
0.584 1 0439
. H m + 00514
.
H 2 m2
Kc ( 46) = K H
K m (1 + K H Km )
0675
.
1 0.472 H m + 0061
. H 2 m2
Kc ( 47) = K H
K m (1 + KH K m )
)(
( 45)
] , T
] , T
( 47)
H K m 0.054 1 + 2.54 H m 0.457 H m
( 46)
K c (45) (1 + K H K m )
2
Kc
( 46)
(1 + K H K m )
H K m 0.0714 1 + 0.685 H m 0131

. H 2 m2
Kc
( 47)
(1 + K H K m )
H K m 00484
.
1 + 143
. H m 0.273 H 2 m 2

( 48)
12
. ( 0.495) H m 1 11
. + 0.698 2
K c (1 + K H K m )
Kc ( 48) = K H
, Ti (48) =
K m (1 + K H K m )
H K m 0.0702 1 + 0.734 ln[ H m ] 1 + 148
. 11
. 2
Kc
( 49)
103
. 1 051
. H m + 0.0759 H 2 m 2
= KH
K m (1 + KH K m )
)(
] , T
0.822 1 0.549 H m + 0112

. H2 m 2
Kc (50) = KH
K m (1 + K H K m )
( 49)
] , T
( 50)
Kc
( 49)
(1 + K H K m )
H K m 0.0386 1 + 3.26 H m 0.6 H 2 m2

=
Kc
( 50)
(1 + K H K m )
H K m 0.0142 1 + 6.96 H m 177

. H 2 m2
Kc
Rule
Ti
Td
Ti (51)
0471
. Ku
K m u
Comment
Servo - nearly min.

IAE, ISE, ITAE Hwang [60]
(continued)
Decay ratio = 0.1 -
Kc(51)
2.4 < 3
Decay ratio = 0.1 -
Kc(5 2)
Model: Method 8
Ti (52)
Kc(53)
Servo nearly
minimum IAE, ITAE
Hwang and Fang
[61]
Model: Method 9

c1 + c2 m + c3 m K u
Tm
Tm
c 1 = 0.537, c 2 = 0.0165
c 3 = 0.00173
0471
. Ku
K m u
3 < 20
Decay ratio = 0.1 -
0471
. Ku
K m u
Ti (53)
01
.
2
Kc

c + c m + c 9 m K u
2
m 7 8 Tm
m
Tm u Kc
K u u c 4 + c 5
+ c6

Tm
Tm
c = 0.350, c = 00344
.
c 4 = 0.0503 , c 5 = 0.163
c 6 = 0.0389
Model: Method 1 or
Method 18
Ku
Am
arbitrary
Ku cos m
Td
Astrom and
Hagglund [98]
Model: Method not
relevant
m
m 1 +
T
+
m
m
0786
.
1 0441
. H m + 00569
.
H 2 m2
Kc (51) = K H
Km (1 + KH K m )
] , T
( 51)
0. 65
m
0 .25 m 1 +
T
+
m
m
0 .65
m
2.0 Tm
decay ratio = 0.03
01
.
Tu
Kc
( 51)
m
2.0 Tm
decay ratio = 0.12
01
.
01
.
Tuning rules
developed from Ku , Tu
4
+ tan 2 m
Specify phase margin

m ; = 4 (Astrom et
al. [30])
(1 + K H K m )
H K m 0.0172 1 + 4.62 H m 0.823 H2 m 2
m
2.0
Tm
Specify gain margin

Am
Tu
4 2 Ti
tan m +
m
2.0 ; 20
Tm
c 9 = 0.00644
2
2
Kc
Simultaneous

c 1 + c 2 m + c 3 m K u
c 7 + c 8 m + c 9 m K u
2
Servo/regulator Tm
Tm
Tm
Tm u K c
K u u c 4 + c 5 m + c 6 m
Tm
Tm
nearly minimum IAE,
c 7 = 0.371, c 8 = 00274
.
c 1 = 0.713, c 2 = 0176
.
ITAE - Hwang and
c 4 = 0.149, c 5 = 0.0556
c 9 = 0.00557
c 3 = 0.0513
Fang [61]
c 6 = 0.00566
Model: Method 9
McMillan [58]
1.415 Tm
1
0 .65
K m m
Tm

1+
Tm + m
m
2.0 ;
Tm
01
.
2
2
Kc
Regulator nearly

m
m
c + c m + c m K u
2
minimum IAE, ITAE c 1 + c 2 T + c 3 T K K u u c 4 + c 5 m + c 6 m 7 8 Tm 9 T m u Kc
m
m
Tm
Tm
Hwang and Fang
c 7 = 0.421, c 8 = 0.00915
c
=
0802
.
,
c
=
0154
.
1
2
[61]
c 4 = 0.190 , c 5 = 0.0532
c 9 = 000152
.
c 3 = 0.0460
Model: Method 9
c 6 = 0.00509
m
2.0 ;
Tm
0.1
( 52)
128
. ( 0.542)
1 0.986 + 0.558 2
K c (1 + K H K m )
Kc (52 ) = K H
, Ti (52) =
K m (1 + K H K m )
H K m 0.0476 1 + 0.996 ln[ H m ] 1 + 2.131 113
. 2
H m
114
. 1 0.466 H m + 0.0647 H 2 m 2
Kc (53) = KH
K m (1 + KH K m )
] , T
)(
( 53)
Kc
( 53)
(1 + K H K m )
H K m 0.0609 1 + 197
. H m 0.323 H 2 m2
Rule
Kc
Ti
Li et al. [99]
Kc(5 4) 5
Td
Model: Method not

relevant
m
0
15
20 0
25 0
Am
2
1.67
1.67
4h
AA m
2.8
3.2
3.5
Tan et al. [39]

Model: Method 10
1.67
1.43
1.43
30
35 0
40 0
3.8
4.0
4.2
45
50 0
55 0
Tu
K u r 2 K
2
4
+ tan 2 m
Ti
2
01443
.
u
04830
.
3.7321
150
0.9330
150
60
65 0
1.11
1.11
5.4
5.5
simplified algorithm
Am = 2 , m = 450 ;
chosen arbitrarily
;
r = 01
. + 09
. (K u K )
05774
.
u
Am
Arbitrary A m , m at
0.25
G p ( j u )
G p j 150 0
4.6
4.9
5.2
00796
.
T
Td
1.25
1.25
1.25
tan m +
rK u
Am
Model: Method not

relevant
Am
Comment
T
4
2
+ +
4

m
Am
03183
.
T
Ku
cos m
Am
Friman and Waller

[41]
Td
m
< 0.25 , Am = 2 ,
Tm
m = 60 0
0.25
m
2.0
Tm
Am = 2 , m = 450
tan b A 2 b2
m
4h cos m
Kc =
, =
with h = amplitude of relay, b =
2b
2
1 + b A2 b 2
2
A m A b +
(Ti Td )
deadband of relay, A, T = limit cycle amplitude and period, respectively.

5
( 54 )
Kc
Rule
Ti
Td
Comment
Direct synthesis
Regulator - Gorecki
et al. [28]
Kc
( 55)
Ti
( 55)
Td
Model: Method 6
Regulator - minimum
IAE - Smith and
Corripio [25] page
343-346
Model: Method 6
Servo minimum
IAE Smith and
Corripio [25] page
343-346
Model: Method 6
Servo 5%
overshoot Smith
and Corripio [25]
page 343-346
Model: Method 6
Kc
( 55)
Ti (56)
Td (5 6)
Tm
K m m
Tm
0.5 m
5Tm
6Km m
Tm
0.5 m
Tm
2K m m
Tm
0.5 m
Kc
(2 tuning rules)
( 56) 6
2
2
m
m m
2 Tm
m
=
6+
3+
9

e
Km m
2Tm
2Tm
2Tm 2Tm
Ti
( 55)
Pole is real and has

maximum attainable
multiplicity; m < 2
Tm
( 5 5)
Low frequency part of

magnitude Bode
diagram is flat.
3+ m 3 m
2 Tm
2Tm
m

m
9 m m
6 +
3+
2Tm
2Tm 2Tm
2Tm
= m
,
2
2
2
3
m
m m
m m
m
21 + 3
+
6
3+
36 4.5

Tm 2Tm
Tm
2Tm
2Tm 2Tm
Td
( 55)
= 0.5 m

3+ m 1
2Tm
2
m

m
9 m m
6 +
3+
2Tm
2Tm 2Tm
2Tm
Kc ( 56) =
1
Km
1
, Ti (56 ) = m
m Tm
2 (56)
+ 1 2
Ti m
1+ 7
Td
( 56 )
= m
T
T
T
Tm
+ 135 m + 240 m + 180 m
m
m
m
m
2
Tm
Tm
Tm
152
+ 1 1 + 3
+ 6
m
m
m

Tm
T
T
T
+ 27 m + 60 m + 60 m
m
m
m
m
2
7 + 42
7 + 42
T
T
T
Tm
+ 135 m + 240 m + 180 m
m
m
m
m
01
.
m
15
.
Tm
01
.
m
15
.
Tm
Suyama [100]
Model: Method 6
1
Tm
+ 0.2236
0.7236
Km
m
Tm + 0.309 m
2.236Tm m
7.236Tm + 2.236 m
Kc
Rule
Juang and Wang
[101]
Model: Method 6
Cluett and Wang
[44]
Model: Method 6
Gain and phase

margin - Zhuang and
Atherton [20]
Model: Method not
relevant
Ti

m
+ 05
. m
Tm
Tm

Km + m
Tm
Model: Method 6
+
Tm
Td

m
+ 05
. m
Tm
Tm
Comment

0.5 m Tm + m 0.5 m
Tm
Tm
Tm
m
m
m
+
+
+ 0.5
Tm
Tm
Tm

+ m
Tm
Closed loop time

constant = Tm ,
0< <1
0.019952 m + 0. 20042Tm
Km m
0.099508 m + 0.99956Tm
m
0.99747 m 8 .7425.10 5 Tm
0.0069905 m + 0.029480Tm
m
0.029773 m + 0.29907 Tm
Closed loop time

constant = 4m
0.055548 m + 0. 33639Tm
Km m
0 .16440 m + 0.99558 Tm
m
0.98607 m 15032
.
.10 4 T m
0.016651 m + 0.093641Tm
m
0.093905m + 0.56867 Tm
Closed loop time

constant = 2m
0.092654 m + 0.43620Tm
Km m
0.20926 m + 0 .98518Tm
m
0.96515 m + 4255010
. 3 Tm
0.024442 m + 0.17669Tm
m
0.17150m + 080740
.
Tm
Closed loop time

constant = 133
. m
012786
.
m + 0.51235Tm
K m m
0.24145 m + 0 .96751Tm
m
0.93566 m + 2 .298810
. 2Tm
0.030407 m + 0 .27480Tm
m
0.25285 m + 1.0132Tm
Closed loop time

constant = m
016051
.
m + 057109
.
Tm
K m m
0 .26502 m + 0 .94291Tm
m
0.89868 m + 6.935510
. 2 Tm
0.035204 m + 0.38823Tm
m
0.33303 m + 11849
.
Tm
Closed loop time

constant = 08
. m
019067
.
m + 0.61593Tm
K m m
0.28242 m + 0.91231T m
m
0.85491 m + 0.15937 Tm
0.039589 m + 0.51941Tm
m
0.40950 m + 13228
.
Tm
Closed loop time

constant = 067
. m
mK u cos( m ),
m = 0 .614(1 0.233 e
0
0 .347 K m K u
0. 45K m Ku
m = 338
. 1 0.97e

0177
.
+ 0.348 m
Tm
Abbas [45]
tan( m ) +
4
+ tan2 ( m )
2 u
tan(m ) +
= 0.413( 3.302K mK u + 1)
1.002
Tm + 05
. m
1 Tm + m
Km Tm m
0.0821
Servo minimum ISE 18578
.
m
- Ho et al. [103]
K
A 0.9087
m
m

Tm
4Tm m
Tm + m
0.9471
Ti
( 57)
V = fractional
overshoot
Tm m
2Tm + m
K m ( 0531
.
0.359V 0.713 )
Camacho et al.[102]
Model: Method 6

margin = 60 0
01
. m 2.0
Tm
4
+ tan2 ( m )
2 u
0 V 0.2
01
. m 5.0
Tm
Tm m
Tm + m
0.4899Tm m 0.1457 m

0.0845
Tm
Am
1.0264
Am [ 2 ,5] ,
m 30 0 ,60 0 ,
m
10
. .
Tm
01
.
Model: Method 6
621189
.
403182
.
m 76.2833 m
+
Am
Tm
A m Tm
(Ho et al [104])
Am [ 2 ,5] ,
Given A m , ISE is minimised when m = 29.7985 +
0.908
0.3678
1.0317
Regulator - minimum 10722
.
m 0.116 m
12497
.
Tm m 1.0082 m
0.4763Tm m 0.328 m
0.2099
0.0961
ISE Ho et al. [103] Km A m 0.8432 Tm
Tm
Tm
Am
Am
Model: Method 6
Given A m , ISE is minimised when m = 46.5489A m0.2035 ( m Tm )
Kc
Rule
Ti
( 57 )
Ti
0.0211Tm 1 + 0.3289 A m + 6.4572 m + 251914

.
( m Tm )
1 + 0.0625A m 0.8079 m + 0.347( m Tm )
Td
0.3693
01
.
m
10
.
Tm
m 30 0 ,60 0 ,
(Ho et al [104])
Comment
Morilla et al. [104a]

Model: Method 24
Kc
Robust
Robust - Brambilla et
al. [48]
Model: Method 6
( 57a ) 8
Ti
1 + 1 4
Tm + 0.5 m
1 Tm + 05
. m
Km
m
= 0. 1; 0 = [ 0. 2, 0. 5]
( 57 a )
m
10 and no
Tm
model uncertainty -
01
.
Tm m
2Tm + m
0.35
1 Tm + 05
. m
Km + 05
. m
Tm + 0.5 m
Tm m
2Tm + m
> 01
. Tm ,
0.8 m .
5Tm
9 m Km
5 m
0.5 m
m
< 033
.
Tm
Model: Method 1
Tm
2 m Km
Tm
0.5 m
m
0.33
Tm
Lee et al. [55]
Ti
K m ( + m )
Rivera et al. [49]

Model: Method 6
Tm +
Model: Method 6
m

1 m
2( + m )
3Ti
m
2( + m )
2
+ m 05
. m
2 + m
2
m
3
3
2
Two degrees of
m
m
freedom
controller;
2
Tm 6
=
T
m
2 + m
T ;
Ti
Tm 1
e
Tm
2
2 + m 05
. m
Desired response =
2 + m
e m s
Tm +
Ti
K m ( 2 + m )
m
m
1 + s
Kc
( 57a )
with 0 =
Ti
( 57 a )
K m [ 2 0 + n 0 m Ti
1
1 +
log e [b a ]
,
2
( 57 a )
)] ,
n0
0 Tm + 0 Tm + Tm m Ti
2
n0
Tm m Ti
( 57 a )
) T
b
= desired closed loop response decay ratio
a
( 57a )
) T
( 57 a ) 2
( 57a ) 2
K m e s
- ideal controller with first order filter
1 + sTm
m
1
1
G c ( s) = Kc 1 +
+ Td s
. 3 tuning rules.
Ti s
Tf s + 1
Rule
Kc
Ti
Td
1 Tm + 05
. m
K m + m
Tm + 0.5 m
Tm m
2Tm + m
TF =
m Tm
m + Tm
Tf =
Comment
Robust
** Morari and
Zafiriou [105]
Horn et al. [106]

Model: Method 6
H optimal Tan et
al. [81]
Model: Method 6
2Tm + m
2 ( + m ) K m
Tm + 0.5 m
0.265 + 0.307 Tm
+ 05
.
Km
m
Tm + 0.5 m
m Tm
m + 2Tm
m
,
2( Tm + m )
> 0.25 m ,
> 0.2Tm .
m
;
2( + m )
> m , < Tm .
m
5.314 + 0.951
= 2 - fast
response
= 1 - robust tuning
= 1.5 - recommended
Tf =
K m e s
- ideal controller with second order filter
1 + sTm
m
1
1 + b1s
G c ( s) = Kc 1 +
+ Td s
. 1 tuning rule.
2
Ti s
1 + a 1s + a 2s
Kc
Ti
Td
2Tm + m
,
2( 2 + m b1) Km
Tm + 05
. m
m Tm
m + 2Tm
Rule
Comment
Robust
Horn et al. [106]
m + 2 Tm m ( m )
2
Model: Method 6
b1 =
Tm ( m + 2 )
2 (2T m )
( m + 2)
Filter
1 + b1 s
1+
2 m + 2 + b1 m
a2 =
2( 2 + m b 1 )
2 m
2( 2 + m b1 )
< Tm
s + a 2s 2
; > m ,
K m e s
- ideal controller with set-point weighting
1 + sTm
m
1
G c ( s) = Kc b +
+ Td s . 1 tuning rule.
Ti s
Kc
Rule
Ti
Direct synthesis
Comment
38
. e
Astrom and
Hagglund [3]
pages 208-210
Td
8.4 + 7.3 2
Tm
K m m
= m ( m + Tm )
52
. me
2.5 1.4 2
0.46Tme 2.8 2.1
or
2
0.89 m e 0.37 4.1 or

2
0.077 Tm e5.0 4 .8
b = 0.40e 0.18 + 2 .8 ;
M s = 1.4 ;
2
014
.
Model: Method 3
m
55
.
Tm
b = 0.22e0.65 + 0.051 ;
Ms = 2.0 ;
2
8.4e 9.6 + 9.8 Tm

Km m
2
32
. m e 1.5 0.93 or
2
0.28Tm e3.8 1.6
0.86 m e 1.9 0.44 or

2
0.076 Tm e3.4 1.1
014
.
m
55
.
Tm
K m e s
- ideal controller with first order filter and set-point
1 + sTm
m
1
1 + 0.4Trs
weighting U(s) = K c 1 +
+ Td s
Y( s) . 1 tuning rule.
R (s )
1 + sTr
Tis
Tf s + 1
Rule
Direct synthesis
Normey-Rico et al.
[106a]
Model: Method not
specified
Kc
Ti
Td
Comment
0. 375( m + 2Tm )
K m m
Tm + 0. 5 m
Tm m
2Tm + m
Tf = 0.13 m
Tr = 0.5 m
K m e s
1 1 + Td s
. 20.
- classical controller G c ( s) = Kc 1 +
1 + sTm
Ti s 1 + Td s
N
m
Kc
Ti
Td
083
. Tm
Km m
15
. m
0.25 m
Rule
Process reaction
Hang et al. [36]
page 76
Model: Method 1
Witt and Waggoner
[107]
Model: Method 1
0.6Tm
Tm
,
]
K m m K m m
Witt and Waggoner

[107]
Model: Method 2
Kc
Regulator tuning
Minimum IAE - Kaya
and Scheib [108]
Model: Method 3
Minimum IAE Witt
and Waggoner [107]
Model: Method 1
Kc
( 58 )
Kc
Ti
( 59 )
aggressive tuning;
05
. Tm
K m m
5 m
0.5 m
conservative tuning;
0. 889Tm
K m m
1.75 m
0.70 m
m
= 0.167
Tm
( 59 )
Tm
098089
.

Km m
0.76167
Tm m

091032
.
Tm
Kc(5 9) 2
Ti (59)
Td (59 )
0.7425 + 0.0150 m + 0.0625 m

Tm
Tm

Tm
+ 0.25 m 0.7425 + 0.0150 m + 0.0625 m

m
Tm
Tm
Tm

Tm
1350
.
+ 025
. 07425
.
+ 0.0150 m + 0.0625 m
m
Tm
Tm
0.921

0.964 Tm m
Tm
1.886
m
1 1 1693
,
.
Tm
1.137

1 1 m 1693
. m
Tm
1.886
, Td

0.59974 Tm m
Tm
1.05221
Tm
m
0718
.
=

Km Tm
Equivalent to Cohen
and Coon [11];
N = [10,20]
0.5 m
2 Km
1350
.
Td
( 5 8)
5 m
T
T
1.350 m + 0.25 m
m
m
Ti (58 ) =
Td (58 ) =
Ti
( 58)
Foxboro EXACT
controller pretune;
N=10
Equivalent to Ziegler
and Nichols [8];
N = [10,20]
Tm
Km m
St. Clair [15] page

21
Model: Method 1
Shinskey [15a]
Model: Method 1
( 5 8) 1
Comment
( 59 )

0.964 Tm m
Tm
1.137

1 + 1 1.693 m
Tm
1.886
0.89819
0<
0.1 <
m
1 ; N=10
Tm
m
< 0.258
Tm
; N = [10,20]
Kc
Rule
Minimum ISE - Kaya
and Scheib [108]
Model: Method 3
Minimum ITAE Kaya and Scheib
[108]
Model: Method 3
Minimum ITAE Witt and Waggoner
[107]
Model: Method 1
Servo tuning
Minimum IAE - Kaya
and Scheib [108]
Model: Method 3
Minimum IAE - Witt
and Waggoner [107]
Model: Method 1
Minimum ISE - Kaya
and Scheib [108]
Model: Method 3
[108]
Model: Method 3
Kc
Td
( 60 )
Kc
Ti
( 60)
( 61)
( 61)
Tm
111907
.

Km m
1.06401
Tm
077902
.

K m m
065
. Tm

Km m
Tm m

07987
.
Tm

0.54766 Tm m
Tm
Tm m

114311
.
Tm
0.70949

0.57137 Tm m
Tm
1.03826
Ti( 60) 3
1.04432

0.762Tm m
Tm
1.03092
Td ( 61)
0.80368
0.86411

0.54568 Tm m
Tm
Tm

0.42844Tm m
Tm
0.99783 + 0.02860 m
Tm
m
1 ; N=10
Tm
0<
m
1 ; N=10
Tm
0<
1.0081
m
1 ; N=10
Tm
0.995
0.914
m
1 1 1392
.
. 013
.
074
,
Tm
Tm

1 m 1 1.392 m
Tm
0.869
m
0.74 0.13
Tm
, Td
( 61)

0.696Tm m
Tm

1 1 1.392 m
Tm
0.869
m
1 ; N = [10,20]
Tm
0<
0.914
m
1 ; N=10
Tm
m
1 ; N=10
Tm
0.995

0.696 Tm m
Tm
; N = [10,20]
0<
1.733
1 1 1.283 m
T ( 60) =
,
1.733
i
Tm
1 1 m 1283
. m
Tm
1.733
m
< 0.379
Tm
0 .1
Tm
112666
.
0.18145 m
Tm
0<
0 .1 <

1.08433
Tm
m
0
.
50814
T
m
0.9895 + 0.09539 m
Tm
Tm

0.762Tm m
Tm

1 + 1 1.283 m
Tm
0.869
Td ( 60)
Ti (61)
Tm
071959
.

Km m
Comment
0.87798
Kc( 61)
0.947
Td
0.9548
Kc( 60)
Tm
112762
.

Km m
0.679 m

K m Tm
m
1086
.
=

Km Tm
Ti
0.89711
0.914
m
0.74 0.13
Tm
Kc
Ti
Td
Kc( 62) 4
Ti (62 )
Td ( 62)
0809
. Tm
Km m
Tm
05
. m
Overshoot = 16%; N=5
Tm
025
. m
N = 2.5
Rule
Minimum ITAE Witt and Waggoner
[107]
Model: Method 1
Direct synthesis
Tsang and Rad [109]
Model: Method 13
aTm
Km m
Tsang et al. [111]
Model: Method 6
1.681
8
1.382
9
0.0 1.161
0
0.1 0.991
6
0.2 0.859
4
0.3 0.754
2
0.4 0.669
3
0.5 0.600
0
0.6 0.542
9
0.7 0.495
7
Comment
0.8 0.456
9
0.9
0 .1
m
1 ; N = [10,20]
Tm
1.0
Robust
Chien [50]
Model: Method 6
1
Tm
K m + 05
. m
Tm
05
. m
= [ m , Tm ] ; N=10
1 05
. m
K m + 05
. m
05
. m
Tm
= [ m , Tm ] ; N=10
K m m
3 m 0.32 Tu
Tu
Tu 015
.
0.05
m
014
. Tu
095
. Tm Km m
1.43 m
052
. m
m Tm = 0.2
095
. Tm Km m
117
. m
048
. m
m Tm = 0.5
114
. Tm Km m
1.03 m
0.40 m
m Tm = 1
139
. Tm Km m
077
. m
0.35 m
m Tm = 2
Ultimate cycle
Minimum IAE
regulator Shinskey
[59] page 167.
Model: Method not
specified
Minimum IAE
regulator - Shinskey
[16] page 143.
Model: Method 6
Kc
Ti
( 62)
( 62)
0.965 m
=

K m Tm
0.85
0.929
m
1 1 1232
.
.
0.796 01465
,
Tm
Tm

0.616 Tm m
Tm

1 m 1 1.232 m
Tm
0.85
0.929
m
0.796 0.1465
Tm
, Td
( 62)

0.616Tm m
Tm

1 1 1.232 m
Tm
0.85
0.929
m
0.796 0.1465
Tm
K m e s
- non-interacting controller
1 + sTm
m
Rule
Regulator tuning
Minimum IAE
Huang et al. [18]
Model: Method 6
Servo tuning
Minimum IAE Huang et al. [18]
Model: Method 6
1
Td s
. 2 tuning rules.
E( s)
U (s) = K c 1 +
Y
(
s
)
Td s
Tis
1+
Kc
Ti
Td
Comment
Kc( 63) 5
Ti (63)
Td ( 63)
01
.
m
1 ; N =10
Tm1
01
.
m
1 ; N =10
Tm
Kc( 64) 6
Ti (64 )
Td ( 64)
0.8058
0.6642
2.1482
m
m
m
1
T
01098
.
8.6290 m + 11863
.
+
231098
.
+
20
.
3519
191463
.
e

Km
Tm
Tm
Tm
Tm
Kc ( 63) =
Ti
( 63)
Td
( 63)
2
3
4
5
= Tm 0.0145 + 2 .0555 m 4.4805 m + 7 .7916 m 7 .0263 m + 2.4311 m

Tm
Tm
Tm
Tm
Tm
2
3
4
5
6
m
m
m
m
m
Tm
m
0.0206 + 0.9385
=
2 .3820 + 7.2774 111018
.
+ 8.0849 2.274
( 63)
Tm
K c
Tm
Tm
Tm
Tm
Tm
0.0865
0.4062
2 .6405
m

m
1
7.0636 + 66.6512 m + 261928

.
+ 7.3453 m
+ 336578
.
57.937e T

Km
Tm
Tm
Tm
Tm
Kc ( 64) =
Ti
( 64 )
Td
( 64 )
2
3
4
5
m

m

= Tm 0.9923 + 0.2819 m 14510

.
.
+ 2.504 m 18759
+ 0.5862 m
Tm
Tm
Tm
Tm
Tm
2
3
4
5
6
m
m
m
m
m
Tm
m
= ( 64) 0.0075 + 0.3449

0.0793 + 0.8089 1.0884 + 0.352 + 0.0471
Tm
K c
Tm
Tm
Tm
Tm
Tm
K m e s
1 + sTm
m
Rule
Tds
1
E (s )
U (s) = K c 1 +
Y (s) . 5 tuning rules.
sT
Tis
1+ d
N
Kc
Ti
Td
Servo tuning
Atherton [20]
Model: Method 6
Minimum ISTSE Zhuang and

Atherton [20]
Model: Method 6
Minimum ISTES Zhuang and

Atherton [20]
Model: Method 6
Tm
1260
.

Km m
0.887
Tm
1295
.

Km m
0.619
Tm
1053
.

Km m
0.930
Tm
1120
.

K m m
0.625
Tm
1001
.

Km m

0.886
Tm
m
0
.
375
T
m
0.701 0.147 m
Tm
Tm

0.378 Tm m
Tm
0.756

0.349 Tm m
Tm
0 .907
Tm

0.350 Tm m
Tm
0.811
Tm

0.308 Tm m
Tm
0.897
Tm

0.308 Tm m
Tm
0.813
Tm
0.661 0.110 m
Tm
Tm
0.736 0.126
m
Tm
0.720 0.114 m
Tm
Tm
0942
.

Km m
0.933
0 .624
Comment
0.770 0.130 m
Tm
0.754 0.116 m
Tm
Ultimate cycle
Servo minimum
.
ISTSE Zhuang and 4.437K m K u 1587
0.037(589
. K m K u + 1)Tu
Ku
Atherton [20]
8.024K m Ku 1435
.
Model: Method not
relevant
Regulator - minimum
0.5556Ku
039
. Tu
IAE - Shinskey [16]
0.4926Ku
0.34Tu
page 148.
05051
.
Ku
0.33Tu
Model: Method 6
0.4608Ku
0.28Tu
0112
. Tu
01
.
m
10
. ; N=10
Tm
11
.
m
2.0 ; N=10
Tm
01
.
m
10
. ; N=10
Tm
11
.
m
2.0 ; N=10
Tm
01
.
m
10
. ; N=10
Tm
11
.
m
2.0 ; N=10
Tm
01
.
m
2.0 ; N=10
Tm
014
. Tu
m Tm = 0.2
014
. Tu
m Tm = 0.5
013
. Tu
m Tm = 1
013
. Tu
m Tm = 2
K m e s
1 + sTm
m
Rule
Td s
1
E (s)
U (s) = K c +
Y ( s) . 6 tuning rules.
sT
Ti s
1+ d
N
Kc
Ti
Td
Regulator tuning
Minimum IAE - Kaya
and Scheib [108]
Model: Method 3
Minimum ISE - Kaya
and Scheib [108]
Model: Method 3
[108]
Model: Method 3
Servo tuning
Minimum IAE - Kaya
and Scheib [108]
Model: Method 3
Tm
113031
.

Km m
Minimum ISE - Kaya

and Scheib [108]
Model: Method 3
Tm
126239
.

Km m
0.8388

[108]
Model: Method 3
Tm
098384
.

Km m
0.49851
Comment
Tm
131509
.

Km m
0.8826
1.3756

0.5655Tm m
Tm
1.25738

0.79715Tm m
Tm
0.41941

0.49547Tm m
Tm
0.41932
Tm m

12587
.
Tm
Tm
13466
.

Km m
0.9308
Tm m

16585
.
Tm
Tm
13176
.

Km m
0.7937
Tm m

112499
.
Tm
0.81314
1.42603
0.4576

0.17707
Tm
m
0.32175Tm
5.7527 5.7241 m
Tm
Tm
Tm

0.47617Tm m
Tm
0.24572
Tm

0.21443Tm m
Tm
0.16768
6.0356 6.0191 m
Tm
2.71348 2.29778 m
Tm
0<
m
1 ; N=10
Tm
0<
m
1 ; N=10
Tm
0<
m
1 ; N=10
Tm
0<
m
1 ; N=10
Tm
0<
m
1 ; N=10
Tm
0<
m
1 ; N=10
Tm
K m e s
- non-interacting controller with set-point weighting
1 + sTm
m
Rule
1
K Ts
U( s) = Kc b +
E( s) c d Y( s) + Kc ( b 1)Y(s) . 3 tuning rules.
sT
Ti s
1+ d
N
Kc
Ti
Td
Comment
Ultimate cycle
Hang and Astrom
[111]
0.6Ku
0.5Tu
0125
. Tu
m
< 0.3 .
Tm
b = 2( x 01
. )+
N=10
Model: Method 1
0.6Ku
0.5Tu
0.6Ku
m
0.5 15
. .083
.
Tu
Tm
0125
. Tu
0.6Ku
m
0.5 15
. .083
.
Tu
Tm
0125
. Tu
0.6Ku
0.335Tu
0125
. Tu
0.5Tu
0125
. Tu
0.6Ku
Hang et al. [65]
Model: Method 1
36
15 K
b=
,
,
'
27 + 5K '
15 + K
10% overshoot - servo 20% overshoot - servo
'
b=
0.6Ku
Hang and Cao [112]

Model: Method 11
K =
11[ m Tm ] + 13
2
37[ m Tm ] 4
0.6Ku

. 0.22 m Tu
053
Tm
'
x = overshoot.
0.3 m < 0.6 .

Tm
b = 2x +
0.6
K =
'
08
.
0125
. Tu
T
. 0.22 m u
053
Tm 4
m
Tm
m
< 10
. ; b=0.8
Tm
10
. <
016
.
0.57
01
.
m
Tm
m
< 0.8 .
Tm
b = 16
.
11[ m Tm ] + 13
2
37[ m Tm ] 4
0.222 K ' Tu
8 4 '
K + 1 ,
17 9
20% overshoot,
10% undershoot;
servo
b=
0125
. Tu
166
. m
,
Tm
m
; b=0.8
Tm
m
< 0.57 ;
Tm
N=10
m
< 0.96 ;
Tm
N=10
m
< 0.5 ; N=10
Tm
K m e s
- industrial controller
1 + sTm
m
1
1 + Td s
U( s) = Kc 1 +
Y(s) . 6 tuning rules.
R( s)
Ts
Ti s
1+ d
N
Rule
Kc
Ti
Regulator tuning
Minimum IAE - Kaya
and Scheib [108]
Model: Method 3
Minimum ISE - Kaya
and Scheib [108]
Model: Method 3
[108]
Model: Method 3
Servo tuning
Minimum IAE - Kaya
and Scheib [108]
Model: Method 3
Tm
081699
.

Km m
Minimum ISE - Kaya

and Scheib [108]
Model: Method 3
Tm
11427
.

Km m
0.9365

[108]
Model: Method 3
Tm
08326
.

Km m
0.7607
Td
Comment
091
. Tm

Km m
0 .7938
Tm m

101495
.
Tm
Tm
11147
.

Km m
0.8992
Tm m

09324
.
Tm
Tm
07058
.

Km m
0.8872
Tm m

103326
.
Tm
1.004
1.00403
0.8753
0.99138

0.5414Tm m
Tm

0.56508 Tm m
Tm
0.7848
0.91107

0.60006Tm m
Tm
0.971

0.97186
Tm
m
0.44278Tm
1.09112 0.22387 m
Tm
Tm
Tm

0.35308Tm m
Tm
0.78088
Tm

0.44243Tm m
Tm
1.11499
0.99223 0.35269 m
Tm
1.00268 + 0.00854 m
Tm
0<
m
1 ; N=10
Tm
0<
m
1 ; N=10
Tm
0<
m
1 ; N=10
Tm
0<
m
1 ; N=10
Tm
0<
m
1 ; N=10
Tm
0<
m
1 ; N=10
Tm
K m e s
1
- series controller Gc ( s) = Kc 1 +
(1 + sTd ) . 3 tuning
1 + sTm
Ti s
rules.
Kc
Ti
Td
Comment
5Tm
6Km m
15
. m
0.25 m
Foxboro EXACT
controller
035
. Ku
0.25
Tu
0.25
Tu
Tm
025
. m
Rule
Autotuning **
Astrom and
Hagglund [3] page
246
Model: Not specified
Ultimate cycle
Pessen [63]
Model: Method 6
01
.
Direct synthesis
aTm
Km m
Tsang et al. [110]

Model: Method 13
a
1.819
4
1.503
9
0.0 1.269
0
0.1 1.089
4
0.2 0.949
2
0.3 0.837
8
0.4 0.748
2
0.5 0.675
6
0.6 0.617
0
0.7 0.570
9
0.8 0.541
3
0.9
1.0
m
1
Tm
K m e s
- series controller with filtered derivative
1 + sTm
m
Table 35: PID tuning rules FOLPD model
1
sTd
Gc ( s) = Kc 1 + 1 +
sT
Ti s
1+ d
. 1 tuning rule.
Kc
Ti
Td
Comment
Chien [50]
Tm
K m ( + 0.5m )
Tm
05
. m
= [ m , Tm ] ; N=10
Model: Method 6
0.5 m
K m ( + 0.5m )
05
. m
Tm
= [ m , Tm ] ; N=10
Rule
Robust
K m e s
- controller with filtered derivative
1 + sTm
m
1
Td s
G c ( s) = Kc 1 +
+
Ti s 1 + s Td
N
Rule
Robust
Chien [50]
Model: Method 6
Gong et al. [113]
. 3 tuning rules.
Kc
Ti
Td
Comment
Tm + 0.5 m
K m ( + 0.5m )
Tm + 05
. m
Tm m
2Tm + m
= [ m , Tm ] ; N=10
Tm + 0.3866 m
K m ( + 1.0009m )
Tm + 0.3866 m
03866
.
Tm m
Tm + 0.3866 m
N = [3,10]
Model: Method 6
( 0.1388 + 0.1247N )Tm + 0.0482 N m

0.3866 ( N 1)Tm + 0.1495 N m
Direct synthesis
Davydov et al. [31]
Model: Method 12
m
Km 1552
.
+ 0078
.
Tm
m
0186
.
+ 0532
. Tm
T
m
K m 1209
.
+ 0103
.
Tm
0.382 m + 0.338 Tm
Tm
m
0.25 0186
.
+ 0532
. Tm
damping factor =
T
m
0.9; 02
. m Tm 1 ;
N = Km
0.4 0.382 m + 0.338 Tm

damping factor =
Tm
0.9; 02
. m Tm 1 ;
N = Km
K m e s
- alternative non-interacting controller 1 1 + sTm
m
1
U( s) = Kc 1 +
E( s) Kc Td sY ( s) . 6 tuning rules.
Ti s
Rule
Kc
Ti
Td
Comment
012
. Tu
Tu
< 2 .7
m
012
. Tu
Tu
2.7
m
Ultimate cycle
2
Ku
Regulator - minimum
IAE - Shinskey [59]
page 167.
Model: method not
specified
Minimum IAE Shinskey [17] page
117. Model: Method
6
143.
Model: Method 6
Regulator - minimum
IAE - Shinskey [16]
page 148.
Model: Method 6
Regulator minimum
IAE - Shinskey [17]
page 121. Model:
method not specified
Process reaction
VanDoren [114]
Model: Method 1
3.73 0.69
Tu
0125
.
Tu
m
0125
.
Tu
m
Ku
T
2.62 0.35 u
m
132
. Tm K m m
180
. m
0.44m
m Tm = 01
.
132
. Tm K m m
1.77m
0.41 m
m Tm = 0.2
135
. Tm Km m
1.43 m
0.41 m
m Tm = 0.5
149
. Tm K m m
117
. m
0.37m
m Tm = 1
182
. Tm Km m
0.92 m
0.32 m
m Tm = 2
0.7692 Ku
0.48Tu
011
. Tu
m Tm = 0.2
0.6993Ku
0.42Tu
012
. Tu
m Tm = 0.5
0.6623Ku
0.38Tu
012T
. u
m Tm = 1
0.6024K u
0.34Tu
012
. Tu
m Tm = 2
0.7576 Ku
0.48Tu
011
. Tu
m Tm = 0.2
15
. Tm
Km m
25
. m
04
. m
K m e s
- Alternative filtered derivative controller 1 + sTm
m
1 1 + 05
. m s + 0.0833 m s 2
G c ( s) = Kc 1 +
. 1 tuning rule.
Ti s
[1 + 01. ms]
Kc
Ti
Td
aTm
Km m
Tm
025
. Tm
Rule
Comment
Direct synthesis
Tsang et al. [110]
Model: Method 13
a
1.851
2
1.552
0
0.0 1.329
3
0.1 1.159
5
0.2 1.028
0
0.3 0.924
6
0.4 0.841
1
0.5 0.768
0
0.6 0.695
3
0.7 0.621
9
0.8 0.552
7
0.9
1.0
K m e s
K
- I-PD controller U(s) = c E (s) K c (1 + Td s)Y(s) . 2 tuning
1 + sTm
Ti s
m
rules.
Kc
Rule
Ti
Td
Comment
Ti ( 64a )
Td ( 64 a )
Underdamped system
Direct synthesis
Chien et al. [74]
K c ( 64a )
m 0.2Tm
Model: Method 6
Minimum ISE
Argelaguet et al.
[114a]. Model:
Method not defined
Kc (64a ) =
Td (64 a ) =
Tm + 0. 5 m
2Tm + m
2K m m
1.414TCLTm + m Tm + 0.25 m 2 TCL2
Km TCL2
+ 0.707TCL m + 0.25 m
, Ti (64 a ) =
0707
. TmTC L m + 0.25Tm m2 05
. m TCL2
2
Tm m + 025
. m + 1414
. TCLTm TC L2
Tm m
2Tm + m
1414
. TCL Tm + mTm + 0.25 m2 TCL2
Tm + 05
. m
First order Pade

approximation for m
Table 40: PID tuning rules - FOLPD model - G m ( s) =
K me s
1 + sTm
m
1
Td s
U (s) = K c 1 +
+
Tis
T
1+ d
Kc
Rule
Ti
Servo/regulator
tuning
Kc
( 64b ) 2
Ti
Kc
Td
( 64 b)
( 64 b)
Td
Comment
Taguchi and Araki

[61a]
T
s
d
E( s) K c +
R (s) . 1 tuning rule.
Td
s
1+
s
1
1. 224
=
0.1415 +
m
Kc
0.001582
Tm
( 64 b)
Td
( 64b )
m
1.0 ;
Tm
20% ; settling time
settling time of
tuning rules of Chien
et al. [10]
2
3
( 64 b)
0.01353 + 2.200 m 1. 452 m + 0. 4824 m
,
T
=
T

m
i
Tm
Tm
Tm
2
2
m
m
m
m
= Tm 0.0002783 + 0. 4119
0. 04943 , = 0. 6656 0. 2786
+ 0. 03966 ,
Tm
Tm
Tm
Tm
= 0. 6816 0.2054
m

+ 0.03936 m
Tm
Tm
1
Table 41: PID tuning rules - non-model specific ideal controller G c ( s) = Kc 1 +
+ Td s . 25 tuning rules.
Ti s
Rule
Kc
Ti
Td
Comment
Ultimate cycle
Ziegler and Nichols
[8]
[ 0.6Ku , Ku ]
0.5Tu
0125
. Tu
Quarter decay ratio
Blickley [115]
0.5K u
Tu
. Tu , 0167
. Tu ]
[ 0125
Quarter decay ratio
0.5K u
Tu
0.2Tu
Parr [64] pages

190-191
05
. Ku
034
. Tu
008
. Tu
Overshoot to servo
response 20%
Quarter decay ratio
De Paor [116]
0.866Ku
0.5Tu
0125
. Tu
phase margin = 30 0
Corripio [117] page

27
Mantz andTacconi
[118]
0. 75Ku
0. 63Tu
01
. Tu
Quarter decay ratio
0.906K u
0.5Tu
0125
. Tu
phase margin = 25 0
0.4698 K u
0.4546Tu
01136
.
Tu
01988
.
Ku
12308
.
Tu
0.3077Tu

margin = 20 0
Gain margin = 2.44,
phase margin = 61 0
0.2015K u
0.7878Tu
01970
.
Tu
Gain margin = 3.45,

phase margin = 46 0
0.35Ku
0.77Tu
0.19Tu
Atkinson and Davey

[119]
** Perry and Chilton
[120]
0. 25Ku
0.75Tu
0. 25Tu
033
. Ku
05
. Tu
033
. Tu
Gain margin 2 , phase

margin 450
20% overshoot - servo
response
Some overshoot
02
. Ku
05
. Tu
033
. Tu
No overshoot
Yu [122] page 11
033
. Ku
05
. Tu
0125
. Tu
Some overshoot
02
. Ku
05
. Tu
0125
. Tu
No overshoot
Luo et al. [121]
0.48Ku
0.5Tu
0125
. Tu
McMillan [14]
page 90
McAvoy and
Johnson [83]
Karaboga and Kalinli
[123]
0.5K u
0.5Tu
0125
. Tu
054
. Ku
Tu
0.2Tu
Astrom and
Hagglund [3] page
142
Astrom and
Hagglund [93]
. Ku ,0.6Ku ]
[ 032
04267
.
K u 15
. Ku
,
Tu
Tu
Tu (1 cos m )
sin m
. Ku Tu ,015
. Ku Tu ]
[ 008
Tu (1 cos m )
4 sin m
m = phase margin
Tu
2
tan m + 1 + tan m
Tu
2
tan m + 1 + tan m
16
small time delay
05
. Ku
12
. Tu
0125
. Tu
aggressive tuning
0.25Ku
12
. Tu
0125
. Tu
conservative tuning
Hang and Astrom

[124]
Ku sin m
Astrom et al. [30]
Ku cos m
St. Clair [15] page 17
Kc
Rule
Pole placement - Shin
et al. [125]
Kc
Other rules
Harriott [126]
pages 179-180
Ti
Td
Ti
( 65)
Comment
Typical : 0.1
Typical : [0.3,0.7]
( 65)
K25%
0167
. T25%
0.667T25%
Quarter decay ratio

Quarter decay ratio
K25%
0.67T25%
017
. T25%
Parr [64] pages

191, 193
05
.
G p ( j u )
Tu
0.25Tu
McMillan [14] Page 43
083
. K25%
0.5T25%
01
. T25%
Fast tuning
0.67K25%
0.5T25%
01
. T25%
Slow tuning
Ti
4
m = 450 ,small m
2 2 G p ( j u )
1
u
Kc( 66) 2
Ti (66 )
Td ( 66)
Calcev and Gorez

[69]
Zhang et al. [127]
Ti
( 65) 1
K c ( 65) =
Ti ( 65) =
1
1
( 65)
( 65)
a 1 uTi
a 2 1Ti
b2 a 2
( 65)
( 65)
u Ti
1Ti
m = 15 0 , large m
( a2 a1 ) + (a 1 a 2 ) 2 + 4(a1 u + b2 1 ) a1 + b 2
2(a1 u + b2 1 )
1
( 1 ) 1 2 , b = 1 sin G ( j ) , a = 1
cos G p ( j 1 ) , = u
2
p
1
1
K1
u
K1
Ku
K1 , 1 = modified ultimate gain and corresponding angular frequency
a2 =
Kc ( 66) =
1 + tan m p
2 A2 2
) + sin
, d, = relay amp. and deadband, A = limit cycle amp.
m
16d 2
Crossing point of the Nyquist curve and relay with hysteresis is outside the unit circle:
2
Ti ( 66) =
u
1
c
u u tan m p
, 10
. +
tan m p 1.2 + u tan m p

c
c
u c tan m p
66
Td ( ) =
, c = frequency when the open loop gain equals unity.
2
2
u c
Crossing point of the Nyquist curve and relay with hysteresis is within the unit circle:
2
Ti ( 66) =
u
1
c
u + u tan p m
u r
r
0.4 u
+ 0.4
r
r
u c tan m p
66
Td ( ) =
, r = oscillation frequency when a pure relay is switched into closed loop.
2
2
u c
Kc
Rule
Garcia and Castelo
[127a]
Kc
Td
( 66a )
( 66a )
Kc
( 66a ) 3
cos 180 0 + m G p ( j1 )
G p ( j1 )
Td
( 66a )
( 66a )
d
Ti
), T
( 66a )
sin 180 0 + m G p ( j1 ) + 1
Ti
21 cos 180 0 + m G p ( j1 )
2[sin 180 0 + m G p ( j1 ) + 1]
Comment
1 cos 180 + m G p ( j1 )
0
) , = oscillation frequency when a sine function is placed in series
with the process in closed loop; 1 < u .
Table 42: PID tuning rules - non-model specific controller with filtered derivative
1
Td s
. 8 tuning rules.
Gc ( s) = Kc 1 +
+
Ti s 1 + Td s
Kc
Ti
Ti
2( A 1 Ti )
A3
Rule
Td
Comment
Td ( 67) 4
N < 10
A3A 4 A2 A 5
2
A 3 A1A 5
N 10
Direct synthesis
A 2 Td
( 67)
Td
A1
Vrancic [72]
A3
2 A 1A 2 A 3K m Td A
2
1
A3
A 2 Td A1
05
. Ti
A1 Km Ti
( 67 ) 2
A2 A 2 2 4 A1A 3
2 A1
Ti
N = 10
= [ 02
. , 0.25]
Td (67 a)
8 N 20
A3
Vrancic [73]
Kc
A3 A 5A1 +
2
Td ( 67) =
(A
2
3
A 5A1
A 2 Td (67 a) A 1
Td (67 a)
Km
N
4
( A 3A 2 A5 )( A5 A2 A4 A 3 )
N
2
( A 3A2 A5 )
N
y( )
y1( t) = Km
d , y2 ( t) =
0
t
y5 ( t) =
( 67a )
[A
(A 1 y1 ())d ,
y3 ( t) =
( A 2 y2 ())d ,
y 4 (t ) =
[A
y 3 ( ) d ,
y 4 ( ) d , A1 = y1( ) , A 2 = y 2 ( ) , A 3 = y3 ( ) , A4 = y 4 ( ) , A 5 = y5 ( )
Alternatively, if the process model is G m ( s) = K m
1 + b1s + b2 s 2 + b3 s3 + b4 s 4 + b5 s5 s
e
, then
1+ a 1s + a 2 s2 + a 3s 3 + a 4 s4 + a 5s 5
m
A1 = Km ( a1 b1 + m ) , A2 = K m b 2 a2 + A1a1 b1 m + 05
. m2 ,
(
(b a
(a b
A3 = Km a 3 b3 + A 2 a1 A1a 2 + b2 m 05
. b1 m 2 + 0167
. m3 ,
A4 = Km
+ A 3a 1 A 2a 2 + A 1a 3 b3 m + 05
. b2 m 2 + 0167
. b1 m 3 + 0.042 m 4 ,
A5 = Km
+ A 4 a1 A 3a 2 + A2 a 3 A1a 4 + b 4 m 05
. b3 m2 + 0167
. b2 m 3 0.042b1 m 4 + 0.008m 5
K c ( 67a ) =
A3
[ T (67 a ) ]2
( 67 a )
2
2 A 1A 2 A 0A 3 Td
A1 d
A 0A 1
N
Td (67a ) is obtained from a solution of the following equation:
A0 A3
( 67a )
Td
N3
A1A 3
( 67 a )
Td
N2
A 0A 5 A 2A 3
( 67 a )
Td
N
] + (A
2
2
3
)[
A 1A5 Td
( 67a )
] + (A A
2
A 3A4 ) = 0
Rule
Kc
Ti
Td
Comment
Direct synthesis
Lennartson and
Kristiansson [157]
Model: Method 1
Kc(111) 5
Ti (111)
0.4Ti (111)
Ku Km 0.6
Kc(111a )
Ti (111a)
0.4Ti (111a)
Ku Km > 10
u K u Km 0.4
Kc(111b )
Ti (111b )
0.4Ti (111b)
Ku K m > 10
u Ku K m 0.4
Kc(111c) 6
Ti (111c)
0.4Ti (111c)
Ku Km > 10
u K u Km 0.4
Kc(111d )
Ti (111d)
0.4Ti (111d )
Ku Km > 10
u Ku K m 0.4
( 111e)
( 111e)
( 111e)
Kristiansson and
Lennartson [157]
Model: Method 1
Kc
12Ku K m 35K u Km + 30
2
Kc (111) = Ku K m
N=
0.4Ti
Ku K m + 25
. 12K u Km 35Ku K m + 30
2
, Ti
( 111)
K c (111)
0.053 u + 0.47 u 014

. u + 0.11
3
2.5
35
30
+
12
Km K u
K m K u K m 2 Ku 2
Kc (111a ) = Ku Km
N=
Ti
Ku K m > 167
.
u Ku K m > 0.45
N = 2.5
2
2
K c (111a )
12 Ku Km 35Ku Km + 30
(111a )
,
T
=
,
i
0.525 u3 + 0.473 u2 0.143 u + 0.113
Ku Km + 3 12Ku 2K m2 35KuK m + 30
3
35
30
+
12
2
2
Km Ku
K mK u Km Ku
2
2
Kc
12Ku Km 35Ku Km + 30
( 111b)
, Ti
=
,
3
2
0185
. u + 1052
. u 0854
. u + 0.309
K uKm + 3 12Ku2 Km 2 35Ku Km + 30
( 111b)
Kc (111b ) = Ku Km
N=
3
Km Ku
35
30
+
12
2
2
K mK u Km Ku
Kc (111c ) = Ku K m
N=
( 111c )
12Ku K m 35K u K m + 30
2
Ku K m + 25
. 12K u Km 35Ku Km + 30
2
, Ti
( 111c )
0525
. u
Kc
,
2
+ 0.473 u 0143
. u + 0113
.
2.5
35
30
+
12
Km K u
K m K u K m 2 Ku 2
Kc (111d ) = Ku K m
12Ku K m 35K u Km + 30
2
( 111d)
Ku K m + 25
. 12K u Km 35Ku K m + 30
2
, Ti
( 111d )
0185
. u
Kc
,
2
+ 1052
. u 0.854 u + 0309
.
2.5
35
30
+
12
Km K u
K m K u K m 2 Ku 2
7.71
914
.
K c (111e )
( 111e )
(111e )
Kc
=
+ 314
. , Ti
=
2
2
3
K m Ku
0.63 u + 0.39 u 2 + 0.15 u + 0.0082
Km K u
N=
Rule
Kc
Ti
Td
Kristiansson and
Lennartson [158]
Model: Method 1
Kc(111f) 7
Ti (111f )
0.4Ti (111f )
Comment
(111g)
Kristiansson and
Lennartson [158a]
Kc
Kc
Model: Method not

specified
Kc (111f ) =
N=
2.5
Km K u
Ti
(111g )
(111g )
Td
(111g )
Tf
(111g)
Kc
Ti
Td
(111h )
Tf
(111h )
Td
(111g )
Td
(111h )
( 111f )
Tf
given below;
0.1 K u K m 0.5
Tf
1. 6( 20 + 13 K m K u )(1+ 0.37 K m K u )
1
2
2
2
K m K u ( 20 + 13K m K u )(1 + 0.37 K m K u ) 2 1.1K m K u 2.3K m K u + 1. 6
2
2
K K (1. 1K u K m 2. 3K u K m + 1.6) 1.6( 20 + 13K m K u )(1 + 0. 37K m K u )
= m u
1
u ( 20 + 13K m K u )(1 + 0. 37K m K u ) 2 1. 1K m 2 K u 2 2. 3K m K u + 1.6
( 20 + 13K m K u ) 2 (1 + 0.37 K m K u ) 2
2
2
2
2
2
K K (1. 1K u K m 2.3K u K m + 1. 6)
(1.1K m K u 2.3K m K u + 1.6)
= m u
u ( 20 + 13K m K u )(1 + 0.37 K m K u ) 2 1.6( 20 + 13K m K u )(1 + 0. 37K m K u )
2
2
1.1K m K u 2. 3K m K u + 1.6
K m K u (1.1K u K m 2.3K u K m + 1.6)
u ( 20 + 13 K m K u )(1 + 0.37 K m K u ) 2
2
2
2
(1.1K u K m 2.3K u K m + 1. 6) 4. 8K m K u (1 + 0.37 K m K u )

1
2
2
2
2
2
3 u K m K u (1+ 0.37 K m K u ) 1. 1K m K u 2. 3K m K u + 1.6
2
2
(1.1K u K m 2.3K u K m + 1.6) 4.8K m K u (1+ 0.37 K m K u )

1
2
2
2
3 u (1 + 0.37 K m K u )
1.1K m K u 2.3K m K u + 1.6
3K m 2 K u 2 (1+ 0.37 K m K u ) 2
2
2
2
2
2
(1.1K u K m 2.3K u K m + 1.6) (1.1K m K u 2.3K m K u + 1. 6)
=
1
2
4
.
8
K
K
(
1
+
0
.
37
K
K
)
3 u (1 + 0. 37K m K u )
m
u
m
u
1
1. 1K m 2 K u 2 2.3K m K u + 1.6
1.1K u K m 2. 3K u K m + 1.6
2
3 u (1 + 0.37 K m K u ) 2
(111h )
given below;
K u K m > 0.5
35
30
+
12
K m K u K m 2 Ku 2
(1. 1K u K m 2. 3K u K m + 1.6)
(111h )
(111h )
Ti
(111h )
(111h ) 8
Ti
(111g )
12 Ku Km 35Ku Km + 30Ku
Kc
Km Ku
, Ti (111f ) =
,
3
3
2
2
2
2
Km Ku + 25
. 12K u Km 35Ku Km + 30
u 095
. Km Ku 2Km Ku + 14
.
Kc
(111g )
Kc
Rule
Kristiansson and
Lennartson [158a]
(continued)
Model: Method not
specified
Other
Kc
Ti
(111i) 9
Ti
Leva [67]
(1
G p ( j )
(1
+ Ti 2
2
Ti Td
+ Ti
2
1+
tan m 0.5
Td (68) 10
Comment
Tf
(111i )
2 tan m + 1 + tan 2 m
u
(111i )
given below;
K u K m < 0.1
= 10
> m
Td ( 68)
6 10
< m
Ti
4
Parameters determined
at m = 300 ,450 ,600
Ku cos m
Astrom [68]
Td
tan m 0.5
Ti Td
Ti
2
(111i )
Ti
G p ( j )
Td
20 .8(1.8 + 0.3K m K135 0 )

1
13 K m (1.8 + 0.3K m K 1350 ) ( 6 + 3. 7K m K 1350 )
( 6 + 3.7K 1350 K m ) K 1350 K m 20. 8(1.8 + 0. 3K m K 1350 )

(111i )
Ti
=
1
131350 (1. 8 + 0.3K m K135 0 ) 2 ( 6 + 3.7K m K 1350 )
2
169 (1. 8 + 0.3K m K135 )
( 6 + 3. 7K m K135 )K m K135
( 6 + 3.7K m K 135 ) K m K 135 ( 6 + 3. 7K m K 135 ) 2
(111i )
( 111i)
Td
=
1 , Tf
=
2 20 .8(1. 8 + 0 .3K K
13 135 (1.8 + 0.3K m K 135 ) 2
13135 (1. 8 + 0.3K m K135 )
m 135 )
1
( 6 + 3.7K m K135 )
Kc
(111i )
( 6 + 3.7K 1350 K m ) 2
10
Td ( 68) =
+ 2 2 + 4 2 tan 2 ( m 05
. )
2 2 tan( m 0.5 )
Table 43: PID tuning rules - non-model specific ideal controller with set-point weighting
1
U( s) = Kc Fp R(s) Y(s) +
( Fi R(s) Y(s)) + Td s( Fd R(s) Y(s)) . 1 tuning rule.
Ts
i
Rule
Kc
Ti
Td
Comment
Ultimate cycle
Mantz and Tacconi
[118]
0.6K u
0.5Tu
Fi = 1
0125
. Tu
Fd = 0.654
Quarter decay ratio
Fp = 017
.
Table 44: PID tuning rules - non-model specific ideal controller with proportional weighting
1
G c ( s) = Kc b +
Ti s
Kc
Rule
Direct synthesis
Astrom and
Hagglund [3] page
217
(0.33e
0.31
Ti
2
)K
= 1 Km Ku
Td
Comment
)T
b = 058
. e 1.3 + 3.5
0 < Km Ku <
maximum Ms = 1.4
)T
b = 025
. e 0.56 0.12
0 < Km Ku <
maximum Ms = 2.0
(0.76e
1.6 0.36
(0.59e
1.3 + 0.38
)T
. e
(017
0.46 2.1
)T
. e
(015
1.4 + 0.56
(0.72e
1.6 + 1.2
)K
Rule
Ultimate cycle
Fu et al. [128]
Table 45: PID tuning rules - non-model specific non-interacting controller
1
K Ts
U( s) = Kc 1 +
E( s) c d Y( s) . 1 tuning rule.
sT
Ti s
1+ d
N
Kc
Ti
Td
0.5K u
0.34Tu
0.08Tu
Comment

1
Table 46: PID tuning rules - non-model specific series controller U( s) = Kc 1 +
(1 + sTd ) . 3 tuning rules.
Ti s
Rule
Kc
Ti
Td
Comment
025
. Ku
033
. Tu
05
.
Tu
optimum servo
response
0.2K u
0. 25Tu
05
.
Tu
0.2K u
0.5Tu
0. 33Tu
0. 33Ku
0. 33Tu
0.5Tu
optimum regulator
response - step
changes
No overshoot; close to
optimum regulator
Some overshoot
025
. Ku
033
. Tu
0.5Tu
Ultimate cycle
Pessen [131]
Pessen [129]
Grabbe et al. [130]
Table 47: PID tuning rules - non-model specific series controller with filtered derivative
1
sTd
. 1 tuning rule.
U( s) = Kc 1 +
1 +
sT
Ti s
1 + d
N
Rule
Kc
Ti
Td
Ultimate cycle
Hang et al. [36]
- page 58
0.35Ku
113
. Tu
0. 20Tu
Comment

1 1 + sTd
Table 48: PID tuning rules - non-model specific classical controller U( s) = Kc 1 +
. 1 tuning rule.
T
Ti s
1+ s d
N
Rule
Kc
Ti
Td
Comment
Ultimate cycle
Corripio [117] page
27
0.6K u
05
. Tu
0125
. Tu
10 N 20
Table 49: PID tuning rules - non-model specific non-interacting controller
1
U( s) = Kc 1 + E ( s) Kc Td sY (s) . 1 tuning rule.
Ti s
Rule
Kc
Ti
Td
0.75K u
0.625Tu
01
. Tu
Ultimate cycle
VanDoren [114]
Comment
Table 50: PID tuning rules - IPD model
K m e s m
s
Kc
Rule
- ideal controller Gc ( s) = K c 1 + 1 + Td s . 5 tuning rules.
Ti s
Ti
Td
Comment
Direct synthesis
Wang and Cluett [76]

K m m
deduced from
Closed loop Damping
graph
time
Factor,
constant
Model: Method 2
m
0.707
Gain margin Phase margin

Am
m [degrees]
2.0
32
0.9056
2.6096
0.3209
2m
0.707
3.1
40
0.5501
4.0116
0.2205
3 m
0.707
4.4
46
0.3950
5.4136
0.1681
4m
0.707
5.5
49
0.3081
6.8156
0.1357
5 m
0.707
6.7
52
0.2526
8.2176
0.1139
6 m
0.707
7.8
54
0.2140
9.6196
0.0980
7m
0.707
8.9
55
0.1856
11.0216
0.0861
8 m
0.707
10.0
56
0.1639
12.4236
0.0767
9m
0.707
11.2
57
0.1467
13.8256
0.0692
10 m
0.707
12.2
58
0.1328
15.2276
0.0630
11m
0.707
13.4
59
0.1213
16.6296
0.0579
12 m
0.707
14.5
59
0.1117
18.0316
0.0535
13 m
0.707
15.6
59
0.1034
19.4336
0.0497
14 m
0.707
16.7
60
0.0963
20.8356
0.0464
15 m
0.707
17.8
60
0.0901
22.2376
0.0436
16 m
0.707
19.0
60
0.0847
23.6396
0.0410
1.0
2.0
37
0.8859
3.2120
0.3541
2m
1.0
2.9
46
0.6109
5.2005
0.2612
3 m
1.0
3.8
52
0.4662
7.1890
0.2069
4m
1.0
4.6
56
0.3770
9.1775
0.1713
5 m
1.0
5.5
58
0.3164
11.1660
0.1462
6 m
1.0
6.4
61
0.2726
13.1545
0.1275
7m
1.0
7.1
62
0.2394
15.1430
0.1311
8 m
1.0
8.0
64
0.2135
17.1315
0.1015
9m
1.0
8.7
65
0.1926
19.1200
0.0921
10 m
1.0
9.5
66
0.1754
21.1085
0.0843
11m
1.0
10.4
67
0.1611
23.0970
0.0777
12 m
1.0
11.1
67
0.1489
25.0855
0.0721
13 m
1.0
12.0
68
0.1384
27.0740
0.0672
14 m
1.0
12.8
68
0.1293
29.0625
0.0630
15 m
1.0
13.4
69
0.1213
31.0510
0.0592
16 m
1.0
14.4
69
0.1143
33.0395
0.0559
Rule
Kc
Ti
Td
Comment
0.9588
Km m
30425
.
m
0.3912 m
Closed loop time

constant = m
Model: Method 2
0.6232
Km m
52586
.
m
0.2632 m
Closed loop time

constant = 2m
0.4668
Km m
7.2291 m
0.2058 m
Closed loop time

constant = 3 m
0.3752
Km m
91925
.
m
01702
.
m
Closed loop time

constant = 4 m
0.3144
Km m
111637
.
m
01453
.
m
Closed loop time

constant = 5 m
0.2709
Km m
131416
.
m
01269
.
m
Closed loop time

constant = 6 m
160
. m
0.48 m
1.48
K mm
2 m
0.37 m
Decay ratio: 2.7:1
094
.
Km m
2 m
0.5 m
Rotach [77]
Model: Method 4
Process reaction
Ford [132]
Model: Method 2
Astrom and
Hagglund [3] page
139
Model: Method not
relevant
121
.
K m m
Damping factor for

oscillations to a
disturbance input =
0.75.
K m e s m
s
- Ideal controller with first order filter, set-point weighting and
1
1 + 0.4Trs
output feedback U(s ) = K c 1 +
+ Tds
Y(s ) K 0 Y(s) . 1 tuning rule.
R( s)
Ti s
1 + sTr
Tf s + 1
Rule
Direct synthesis
Normey-Rico et al.
[106a]
Model: Method not
specified
Kc
Ti
Td
Comment
Tf = 0. 13m
0.563
K m m
1.5m
0.667 m
1
2Km m
Tr = 0.75 m
K0 =
K m e s m
s
- ideal controller with filtered derivative
1
Td s
G c ( s) = Kc 1 +
+
sT
Ts
i
1+ d
Rule
Robust
Chien [50]
Model: Method 2
. 1 tuning rule.
Kc
Ti
Td
2
K m ( + 0.5m )
2 + m
m ( + 025
. m )
2 + m
Comment
1
; N=10
Km
K m e s m
- series controller with filtered derivative
s
1
Td s
. 1 tuning rule.
G c ( s) = Kc 1 + 1 +
Ts
Ts
i
1 + d
Rule
Kc
Ti
Td
Comment
1 2 + 0.5 m
2
K m [ + 05
. m ]
2 + 05
. m
05
. m
=
, m ; N=10
K m
1
05
. m
2
K m [ + 05
. m ]
05 m
2 + 05
. m
=
, m ; N=10
K m
Robust
Chien [50]
Model: Method 2
1 1 + Td s
.
Ti s 1 + Td s
N
5 tuning rules.
K e s m
Table 54: PID tuning rules - IPD model m
s
Rule
Kc
Ti
Td
Comment
Ultimate cycle
Luyben [133]
Model: Method 2
0.46Ku
2.2Tu
016
. Tu
maximum closed loop

log modulus of +2dB ;
N=10
311
. Ku
2.2Tu
364T
. u
N=10
0.5m
2 + 05
. m
=
, m ; N=10
K
m
2 + 0.5 m
0.5m
=
, m ; N=10
K m
Belanger and
Luyben [134]
Model: Method 2
Robust
Chien [50]
Model: Method 2
Regulator tuning
121. Model: Method
not specified
117. Model: Method
1
Minimum IAE
Shinskey [59] page
74
Model: Method not
specified
1
Km
05
. m
+ 0.5 2
[
m]
1 2 + 0.5 m
2
K m [ + 05
. m ]
056
. Ku
0.39Tu
015
. Tu
0.93
Km m
157
. m
056
. m
09259
.
Km m
160
. m
0.58 m
N=10
09259
.
Km m
1.48 m
0.63m
N=20
K m e s m
s
- Alternative non-interacting controller 1 -
1
U( s) = Kc 1 +
E( s) Kc Td sY ( s) . 2 tuning rules.
Ti s
Rule
Regulator tuning
74.
Model: Method not
specified
Minimum IAE
Shinskey [17] page
121. Model: Method
not specified
Minimum IAE
Shinskey [17] page
117. Model: Method
1
Kc
Ti
Td
12821
.
K m m
19
. m
046
. m
0. 77K u
0. 48Tu
015
. m
128
.
K m m
1.90 m
0.48 m
Comment
Kc
Rule
Direct synthesis
Chien et al. [74]
Model: Method 1
Kc
( 68a )
Kc
( 68a ) 1
1.414 TCL + m
2
CL
Km T
+ 0.707TCL m + 0.25 m
K m e s m
- I-PD controller U(s) = Kc E (s) K c(1 + Td s)Y(s) .
s
Ti s
1 tuning rule.
Ti
Td
Comment
1414
. TCL + m
0.25 m2 + 0.707TC L m
1.414TCL + m
Underdamped system
m 0.2Tm
Table 57: PID tuning rules - IPD model G m ( s) =
U(s) = Kc (1 +
Rule
Direct synthesis
Hansen [91a]
Model: Method not
specified
K me s m
s
- controller
1
) E( s) + Kc ( b 1) R (s) Kc TdsY( s) . 1 tuning rule.
Tis
Kc
Ti
Td
Comment
0.938 K m m
2. 7m
0.313 m
b = 0.167
K m e s m
s
1
Td s
U (s) = K c 1 +
+
Tis
T
1+ d
Rule
Kc
Servo/regulator
tuning
1 1.253
K m m
= 0. 6642
c = phase
corresponding to the
crossover frequency;
Km =1
Td
Comment
Taguchi and Araki

[61a]
Minimum ITAE Pecharroman and

Pagola [134a]
Ti
T
s
d
E( s) K c +
R (s) . 2 tuning rules.
Td
s
1+
s
2.388 m
0.4137 m
= 0.6797
1.672 K u
0.366 Tu
0.136 Tu
1.236 K u
0.427 Tu
0.149 Tu
0.994 Ku
0.486 Tu
0.155 Tu
0.842 Ku
0.538 Tu
0.154 Tu
0.752 Ku
0.567 Tu
0.157 Tu
0.679 Ku
0.610 Tu
0.149 Tu
0.635 K u
0.637 Tu
0.142 Tu
0.590 Ku
0.669 Tu
0.133 Tu
0.551K u
0.690 Tu
0.114 Tu
0.520 K u
0.776 Tu
0.087 Tu
0.509 Ku
0.810 Tu
0.068 Tu
m
1.0 ;
Tm
20% ; settling time
settling time of
et al. [10]
= 0. 601 , = 1 ,
N = 10, c = 1640
= 0.607 , = 1 ,
N = 10, c = 1600
= 0.610 , = 1 ,
N = 10, c = 1550
= 0.616 , = 1 ,
Model: Method 6
N = 10, c = 1500
= 0.605 , = 1 ,
N = 10, c = 1450
= 0.610 , = 1 ,
N = 10, c = 1400
= 0.612 , = 1 ,
N = 10, c = 1350
= 0.610 , = 1 ,
N = 10, c = 1300
= 0.616 , = 1 ,
N = 10, c = 1250
= 0.609 , = 1 ,
N = 10, c = 1200
= 0. 611 , = 1 ,
N = 10, c = 1180
Km e s
- ideal controller Gc ( s) = K c 1 + 1 + Td s .
s(1 + sTm )
Ti s
Table 59: PID tuning rules - FOLIPD model
1 tuning rule.
Rule
Kc
Ti
Td
Comment
Ultimate cycle
McMillan [58]
Model: Method not
relevant
Kc( 69) 1
Ti ( 69 )
Td ( 69)
Tuning rules
Kc ( 69)
T 0.65
T 0.65
1111
.
Tm
1
( 69 )
( 69 )
m
=
. m 1 + m
, Ti = 2 m 1 + , Td = 05
K m m 2 T 0.65
m
m
m
1 +
m
Km e s
- ideal controller with filter
s(1 + sTm )
m
1
1
G c ( s) = Kc 1 +
+ Td s
. 3 tuning rules.
Ti s
1 + Tf s
Kc
Ti
Td
0.463 + 0277
.
[
]
2
Km m
m
Tm +
0.238 + 0123
.
Tm m
0238
.
+
0123
. ) Tm + m
(
Rule
Comment
Robust
Tan et al. [81]

Model: Method 2
Zhang et al. [135]
Model: Method 2
Tan et al. [136]

Model: Method 2
([ 0238
. + 0123
. ]Tm + m )
3 + Tm + m
K m 3
+ 3 m + m2
3 + Tm + m
( 3 + m ) Tm
3 + m + Tm
= 15
. m .Overshoot = 58%, Settling time = 6 m
= 25
. m .Overshoot = 35%, Settling time = 11 m
= 35
. m .Overshoot = 26%, Settling time = 16m
= 45
. m .Overshoot = 22%, Settling time = 20m
m
5.750 + 0.590
= 0.5 - performance
= 0.1 - robustness
= 0.25 - acceptable
Tf =
Tf =
3
3 2 + 3 m + m 2
15
. m 4.5 m
Obtained from graph
Tm m
0. 0337Tm
m
1 +
2
01225
.
T
Km m
m
Tm + 81633
.
m
01225
.
Tm + m
0. 0754Tm
m
1+
2
0
.
1863
T
Km m
m
Tm + 53677
.
m
01863
.
Tm + m
01344
.
Tm
m
1+
0.2523Tm
K m m2
Tm + 3.9635 m
0.2523Tm + m
Tf = 05549
.
m
Tm m
Tf = 0. 4482 m
Tm m
Tf = 0. 2863m
Km e s
s(1 + sTm )
m
1
G c ( s) = Kc b +
Ti s
Rule
Kc
Ti
Td
Comment
Direct synthesis
Astrom and
Hagglund [3] - pages
212-213
56
. e 8 .8 + 6.8
,
K m ( Tm + m )
11
. m e 6.7 4.4
= m ( m + Tm )
Model: Method 1 or
Method 2.
86
. e 7.1 + 5.4
K m ( Tm + m )
10
. m e3.3 2 .33
17
. m e 6.4 + 2.0
Maximum Ms = 1.4
b = 012
. e 6.9 6.6
2
0.38 me 0.056 0.60
Maximum Ms = 2.0
b = 056
. e 2 .2 + 1.2
Km e s
1 1 + Td s
.
s(1 + sTm )
Ti s 1 + Td s
N
5 tuning rules.
m
Kc
Rule
Ti
Td
Comment
Tm
2 + m
= [ m , Tm ] ; N=10
2 + m
Tm
= [ m , Tm ] ; N=10
Robust
Chien [50]
Model: Method 2
Regulator tuning
Minimum IAE
Shinskey [59] page
75.
Model: Method not
specified
1
Km
+ 2
[
]
m
1 2 + m
K m [ + m ] 2
1.38( m + Tm )
0. 78
K m ( m + Tm )
100
Minimum IAE 108K m m 122

. 0.03
m
Shinskey [59]
pages 158-159
100
Model: Open loop
T
108K m m 1 + 0. 4 m
m
Tm

Pomerleau [82], [92]
deduced from graph
Model: Method 2
Output step load

disturbance
Input step load

disturbance
K m ( m + Tm )
Tm 2
a ( m + T m )
0.66( m + Tm )
m = Tm ; N=10
157
. m 1 + 12
. 1 e m
0.56 m + 0.75Tm
Tm
> 05
.
m
157
. m 1 + 12
. 1 e m
0.56 m + 075
. Tm
Tm
05
.
m
a( m + Tm )
+1
Tm
m
2;
( Td N)
01
. Tm
Td
033
. Tm
N
m ( Td N)
m ( Td N)
0.2
0.4
0.6
0.8
1.0
5.0728
4.9688
4.8983
4.8218
4.7839
0.5231
0.5237
0.5241
0.5245
0.5249
1.2
1.4
1.6
1.8
2.0
4.7565
4.7293
4.7107
4.6837
4.6669
0.5250
0.5252
0.5254
0.5256
0.5257
0.2
0.4
0.6
0.8
1.0
3.9465
3.9981
4.0397
4.0397
4.0397
0.5320
0.5315
0.5311
0.5311
0.5311
1.2
1.4
1.6
1.8
2.0
4.0397
4.0278
4.0278
4.0218
4.0099
0.5312
0.5312
0.5312
0.5313
0.5314
Km e s
- series controller with derivative filtering
s(1 + sTm )
m
1
Td s
. 1 tuning rule.
G c ( s) = Kc 1 + 1 +
Ts
Ts
i
1 + d
N
Kc
Rule
Ti
Td
2 + m
Tm
Tm
2 + m
Robust
Chien [50]
Model: Method 2
1
Km
2 +
m
+ 2
[
m]
1
Tm
2
K m [ + m ]
Comment
= [ m , Tm ] ; N=10
(Chien and Fruehauf
[137])
= [ m , Tm ] ; N=10
Km e s
- alternative non-interacting controller 1
s(1 + sTm )
m
1
U(s) = Kc 1 + E (s) K c Td sY ( s) . 2 tuning rules.
Ti s
Rule
Regulator tuning
75.
Model: Method not
specified
159.
Model: Open loop
method not defined
Kc
Ti
Td
118
.
K m ( m + Tm )
1.28
K m m (1 + 0.24
Tm
m
T
014
. m )
m
1.38( m + Tm )
0.55( m + Tm )
Tm
19
. m 1 + 0.75[1 e m ]
0.48 m + 0.7Tm
Comment
Km e s
- ideal controller with filtered derivative
s(1 + sTm )
m
1
Td s
. 1 tuning rule.
G c ( s) = Kc 1 +
+
Ti s 1 + s Td
N
Rule
Kc
Robust
Chien [50]
2 + m + Tm
Model: Method 2
K m( + m)
Ti
2 + Tm + m
Td
Tm ( 2 + m )
2 + Tm + m
Comment
= Tm ; N = 10
Km e s
s(1 + sTm )
m
U(s) = Kc Fp R ( s) Y( s) +
Kc
[ Fi R(s) Y(s)] + Kc Td s[ Fd R (s) Y(s)] . 1 tuning rule.
Ti s
Rule
Kc
Ti
Td
Ultimate cycle
Oubrahim and
Leonard [138]
Model: Method not
relevant
06
. Ku
Fp = 01
.
05
. Tu
Fi = 1
0125
. Tu
Fd = 0.01
Comment
m
< 0.8 ;
Tm
20% overshoot
0.05 <
Km e s
- Alternative classical controller
s(1 + sTm )
m
1+ T s 1 + T s
i
d
G c ( s) = Kc
. 1 tuning rule.
1 + Td s 1 + Td s
N
N
Rule
Direct synthesis
Tsang and Rad [109]
Model: Method 3
Kc
Ti
Td
Comment
0.809
K m m
Tm
0.5 m
Maximum overshoot =
16%; N = 8.33
Km e s
s(1 + sTm )
m
Table 68: PID tuning rules FOLIPD model
1
Td s
U (s) = K c 1 +
+
Tis
T
1+ d
Rule
Kc
Servo/regulator
tuning
Taguchi and Araki
[61a]
Ti
T
s
d
E( s) K c +
Td
s
1+
s
Td
Kc
( 69a ) 2
Ti
( 69a )
Td
( 69a )

Pagola [134a]
c = phase
Km = 1 ; Tm = 1 ;
0.05 < m < 0. 8 .
Comment
1.672 K u
0.366 Tu
0.136 Tu
1.236 K u
0.427 Tu
0.149 Tu
0.994 Ku
0.486 Tu
0.155 Tu
0.842 Ku
0.538 Tu
0.154 Tu
0.752 Ku
0.567 Tu
0.157 Tu
0.679 Ku
0.610 Tu
0.149 Tu
0.635 K u
0.637 Tu
0.142 Tu
0.590 Ku
0.669 Tu
0.133 Tu
0.551K u
0.690 Tu
0.114 Tu
m
1.0 ;
Tm
20% ; settling time
settling time of
et al. [10]
= 0. 601 , = 1 ,
N = 10, c = 1640
= 0.607 , = 1 ,
N = 10, c = 1600
= 0.610 , = 1 ,
N = 10, c = 1550
= 0.616 , = 1 ,
N = 10, c = 1500
= 0.605 , = 1 ,
Model: Method 4
N = 10, c = 1450
= 0.610 , = 1 ,
N = 10, c = 1400
= 0.612 , = 1 ,
N = 10, c = 1350
= 0.610 , = 1 ,
N = 10, c = 1300
= 0.616 , = 1 ,
Kc
Td
( 69a )
( 69 a )
1
0.5184
=
0.7608 +
m
Kc
[
0.01308 ] 2
Tm
( 69 a )
0.03330 + 3.997 m 0.5517 m
,
T
=
T

m
i
Tm
Tm
2
3
m
m
m
= Tm 0.03432 + 2. 058
1. 774 + 0.6878 ,
Tm
Tm
Tm
= 0. 6647 , = 0. 8653 0. 1277
N = 10, c = 1250
m

+ 0.03330 m
Tm
Tm
Rule
Pagola [134a] continued
Kc
Ti
Td
0.520 K u
0.776 Tu
0.087 Tu
0.509 Ku
0.810 Tu
0.068 Tu
Comment
= 0.609 , = 1 ,
N = 10, c = 1200
= 0. 611 , = 1 ,
N = 10, c = 1180
Km e s
K m e s
or
- ideal controller
(1 + sTm1 )(1 + sTm 2 ) Tm12 s2 + 2 m Tm1s + 1
m
Table 69: PID tuning rules - SOSPD model
1
Gc ( s) = K c 1 +
Ti s
Kc
Rule
Servo tuning
Minimum ITAE
Sung et al. [139]
Model: Method 2
Regulator tuning
Minimum ITAE
Sung et al. [139]
Model: Method 2
Ti
Td
Comment
Kc( 70)
Ti (70 )
Td
( 70)
0.05 <
m
2
Tm1
0.05 <
m
2
Tm1
Kc( 71)
Ti (71)
Td
( 71)
Kc
Kc
( 70)
( 70)
0.983

1
0.04 + 0333
m , m 0.9 or
.
+ 0.949 m
Km
Tm1
0.832
m
1
m
=
0.544 + 0308
.
+ 1408
.
m , m > 0.9 .
Km
Tm1
Tm1

Ti ( 70) = Tm1 2.055 + 0.072 m m , m 1 or Ti ( 70) = Tm1 1768
.
+ 0.329 m m , m > 1
T
T
T
T
m1
m1
m1
m1
Td ( 70) =
0.870
1 e
1.060
m1
Kc
( 71)
Tm1
1.090
T
0.55 + 1.683 m1
m

2.001
0.766

m
m
1
=
067
. + 0297
.
+ 2189
.
m , m < 0.9 or
Km
Tm1
Tm1
Tm1
Kc ( 71) =
2
0.766

1
0365
.
+ 0.260 m 14
. + 2189
. m
m , m 0.9
Km
Tm1
Tm1
Tm1

Ti ( 71) = Tm1 2.212 m
Tm1
Ti
( 71)
0.520

03
. , m < 0.4 or
Tm1
0.15+0.33 m
m
Tm1
= Tm1{0.975 + 0.910
1.845 + 1 e
Tm1
Td ( 71) =
} , m 0.4
5
.
25
0
.
88
2
.
8
Tm1
Tm1
Tm1
m
1.121

1.171
0.530
0.15+ 0.939 m
T

Tm 1
1 e
145
. + 0.969 m1 19
. + 1576
. m
m
Tm1
Kc
Rule
Minimum ITAE
Lopez et al. [84]
- taken from plots
Ti
Td
Comment
m Tm1
Kc
Ti
Td
0.5
0.1
25
Km
05
. Tm1
0.25Tm1
0.5
1.0
07
.
Km
1.3Tm1
12
. Tm1
0.5
10.0
0.35
Km
5Tm1
1.0Tm1
1.0
0.1
25
Km
0.5Tm1
0.2Tm1
1.0
1.0
18
.
Km
1.7Tm1
0.7Tm1
4.0
1.0
9.0
Km
2Tm1
0.45Tm1
Model: Method 12
Representative results
Ultimate cycle
Regulator nearly
minimum IAE, ISE,
ITAE Hwang [60]
Kc
( 72) 2
Ti
( 72 )
1.45(1 + Ku K m )
116
.
1
2
0
Km u
Decay ratio = 0.15

0.2 m 2.0 and
Tm1
. u2 m2 )
( 1 0.612 u m + 0103
0.6 m 4.2 ;
Model: Method 3
Kc
( 73)
Ti
(73)
1.45(1 + Ku K m )
116
.
1
2
0
Km u
. u2 m2 )
( 1 0.612 u m + 0103
< 2.4
Decay ratio = 0.15

0.2 m 2.0 and
Tm1
0.6 m 4.2 ;
2.4 < 3
KH =
m 2
T
4 K mTd Kc m
9
2
Tm1 m m m +
+
2
9
9 u
2 K m m 18
4
2
2
2
2
2
3
3
3
2
2 2
2
9
T 4 + m + 49 m m Tm m Tm1 + 7Tm m m + 10 m1 Tm1 m K T K 10 m + 4 Tm1 m ( m m + Tm1 ) 8Kc Km Td m
m1
c d m
2
2
324
81
9
81
9
81
9
2 Km m
81 u
H =
Tm12
Kc
1 + K HK m
6T 2 + 4Tm1 m m + K uK m m 2
, 0 = m1
2
2 m Tm12 u
2 T
K K
2 Kc KmTd m
+ m m m+ H m m
3
6
3 u
6Tm12 + 4 Tm1m m + KH Km m 2 4 mK mK cTd

( Km Kc Td m 2 + 2 mTm12 ) H
( 72)
] , T
] , T
0674
.
1 0.447 H m + 0.0607 H2 m 2
= KH
K m (1 + KH K m )
0.778 1 0.467 H m + 0.0609 H 2 m 2

Kc ( 73) = KH
K m (1 + K H K m )
( 72)
( 73)
Kc
( 72)
(1 + K H K m )
H K m 00607
.
1 + 105
. H m 0.233 H 2 m 2
Kc
( 73)
(1 + K H K m )
H K m 0.0309 1 + 2.84 H m 0.532 H 2 m 2
Rule
Kc
Ti
Td
Comment
Regulator nearly
minimum IAE, ISE,
ITAE Hwang [60]
continued
Kc( 74) 3
Ti (74 )
1.45(1 + Ku K m )
116
.
1
2
0
Km u
Decay ratio = 0.15

0.2 m 2.0 and
Tm1
. u2 m2 )
( 1 0.612 u m + 0103
0.6 m 4.2 ;
3 < 20
Model: Method 3
Kc( 75)
Decay ratio = 0.15

0.2 m 2.0 and
Tm1
1.45(1 + Ku K m )
116
.
1
2
0
Km u
Ti (75)
. u2 m2 )
( 1 0.612 u m + 0103
0.6 m 4.2 ; 20
Kc( 76)
Decay ratio = 0.2

0.2 m 2.0 and
Tm1
1.45(1 + Ku K m )
116
.
1
2
0
Km u
Ti (76 )
. u2 m2 )
( 1 0.612 u m + 0103
0.6 m 4.2 ; < 2.4
Kc
( 77)
Ti
1.45(1 + Ku K m )
116
.
1
2
0
Km u
( 77 )
. u2 m2 )
( 1 0.612 u m + 0103
Decay ratio = 0.2

0.2 m 2.0 and
Tm1
0.6 m 4.2 ;
2.4 < 3
Kc( 78)
1.45(1 + Ku K m )
116
.
1
2
0
Km u
Ti (78)
. u2 m2 )
( 1 0.612 u m + 0103
Decay ratio = 0.2

0.2 m 2.0 and
Tm1
0.6 m 4.2 ;
3 < 20
Kc( 79)
Decay ratio = 0.2

0.2 m 2.0 and
Tm1
1.45(1 + Ku K m )
116
.
1
2
0
Km u
Ti (79 )
. u2 m2 )
( 1 0.612 u m + 0103
0.6 m 4.2 ; 20
( 74)
131
. ( 0519
.
. + 0514
.
2
) H m 1 103
K c (1 + K H K m )
Kc ( 74) = K H
, Ti (74) =
K m (1 + KH K m )
H K m 0.0603 1 + 0.929 ln[ H m ] 1 + 2.01 12
. 2
114
. 1 0.482 H m + 0.068 H 2 m 2
Kc ( 75) = KH
K m (1 + K H K m )
Kc
( 76)
Kc
( 77)
Kc
( 78)
] , T
0.622 1 0.435 H m + 0.052 H m
= KH
K m (1 + K H K m )
( 75)
] , T
0.724 1 0.469 H m + 0.0609 H 2 m 2
= KH
K m (1 + K H K m )
)(
( 76)
] , T
Kc
( 75)
(1 + K H K m )
H K m 0.0694 1 + 21
. H m 0.367 H2 m 2
( 77)
K c (76) (1 + K H Km )
H Km 0.0697 1 + 0.752 H m 0145

. H 2 m2
=
Kc
( 77 )
(1 + KH K m )
H K m 00405
.
1 + 193
. H m 0363
. H 2 m2

( 78)
126
. ( 0.506) H m 1 1.07 + 0.616 2
K c (1 + K H K m )
= KH
, Ti (78) =
K m (1 + KH K m )
H K m 0.0661 1 + 0.824 ln[ H m ] 1 + 171
. 117
. 2
109
. 1 0497
. H m + 00724
.
H 2 m2
Kc ( 79) = K H
Km (1 + KH K m )
)(
] , T
( 79)
Kc
( 79 )
(1 + KH K m )
H K m 0.054 1 + 2.54 H m 0457

. H 2 m2
Rule
Kc
Ti
Td
Regulator nearly
minimum IAE, ISE,
ITAE Hwang [60]
continued
Kc(8 0) 4
Ti (80)
Comment
1.45(1 + Ku K m )
116
.
1
2
0
Km u
. u2 m2 )
( 1 0.612 u m + 0103
Decay ratio = 0.25

0.2 m 2.0 and
Tm1
0.6 m 4.2 ; < 2.4
Kc
Model: Method 3
( 81)
Ti
1.45(1 + Ku K m )
116
.
1
2
0
Km u
( 81)
. u2 m2 )
( 1 0.612 u m + 0103
Decay ratio = 0.25

0.2 m 2.0 and
Tm1
0.6 m 4.2 ;
2.4 < 3
Kc(8 2)
1.45(1 + Ku K m )
116
.
1
2
0
Km u
Ti (82)
. u2 m2 )
( 1 0.612 u m + 0103
Decay ratio = 0.25

0.2 m 2.0 and
Tm1
0.6 m 4.2 ;
3 < 20
Kc(83)
1.45(1 + Ku K m )
116
.
1
2
0
Km u
Ti (83)
Decay ratio = 0.25

0.2 m 2.0 and
Tm1
. u2 m2 )
( 1 0.612 u m + 0103
0.6 m 4.2 ; 20
Servo nearly
minimum IAE, ISE,
ITAE Hwang [60].
In all, decay ratio =
0.1 with
0.2 m 2.0 and

Tm1
Kc(8 4)
Ti (84 )
0471
. Ku
K m u
Kc(85)
Ti (85)
0471
. Ku
K m u
m
+ 0117
. m
Tm1
Tm1
0613
.
+ 0613
.

m
+ 0117
. m
Tm1
Tm1
0613
.
+ 0613
.
0.6 m 4.2
Model: Method 3
Kc
( 80)
0.584 1 0.439 H m + 0.0514 H 2 m 2
= KH
K m (1 + K H K m )
( 80)
] , T
0675
.
1 0.472 H m + 0061
. H 2 m2
Kc (81) = K H
K m (1 + KH K m )
] , T
( 81)
Kc
(1 + KH K m )
H K m 0.0714 1 + 0.685 H m 0.131 H 2 m 2

Kc
( 80)
( 81)
(1 + K H K m )
H K m 00484
.
1 + 143
. H m 0.273 H 2 m 2

( 82)
12
. ( 0.495) H m 1 11
. + 0.698 2
K c (1 + K H K m )
Kc (82 ) = K H
, Ti (82) =
K m (1 + K H K m )
H K m 0.0702 1 + 0.734 ln[ H m ] 1 + 148
. 11
. 2
103
. 1 0.51 H m + 0.0759 H 2 m 2
Kc (83) = KH
K m (1 + K H K m )
] , T
0.822 1 0.549 H m + 0112

. H m
Kc (84 ) = K H
K m (1 + K H K m )
)(
( 83)
] , T
0.786 1 0.441 H m + 0.0569 H 2 m 2

Kc (85) = KH
K m (1 + K H K m )
( 84)
] , T
Kc
( 83)
(1 + KH K m )
H K m 0.0386 1 + 3.26 H m 0.6 H 2 m 2

=
( 85)
Kc
( 84 )
(1 + K H K m )
H K m 0.0142 1 + 6.96 H m 177

. H 2 m2
Kc
( 85)
(1 + K H Km )
H K m 0.0172 1 + 4.62 H m 0.823 H 2 m 2
Kc
Rule
Servo nearly
minimum IAE, ISE,
ITAE Hwang [60]
(continued)
Kc
Servo nearly
minimum IAE, ISE,
ITAE Hwang [60]
Ti
( 8 6) 5
Ti
Td
Ti (87 )
( 8 8)
( 88)
Kc
Ti
0471
. Ku
K m u
( 86)
Kc(8 7)
0471
. Ku
K m u
0471
. Ku
K m u
Ti (89)
0471
. Ku
K m u
Kc( 90)
Ti (90)
0471
. Ku
K m u
Kc
( 91)
Ti
( 91)
0471
. Ku
K m u
Kc
( 92)
Ti
( 92 )
0471
. Ku
K m u
Kc(8 9)
Comment
m
+ 0117
. m
Tm1
Tm1
0613
.
+ 0613
.

m
+ 0117
. m
Tm1
Tm1
0613
.
+ 0613
.
> 0649
.
+ 058
.
m
Tm1
> 0649
.
+ 058
.
Model: Method 3
Kc( 93)
0649
.
+ 058
.
m
Tm1
m
Tm1
m
Tm1
0005
.
Tm1
0005
.
Tm1
2
m
0005
. m
Tm1
Tm1
> 0613
.
+ 0613
.
0649
.
+ 058
.
2
m
0005
. m
Tm1
Tm1
> 0613
.
+ 0613
.
Kc
( 88)
] , T
0.794 1 0.541 H m + 0126

. H 2 m2
= KH
K m (1 + K H K m )
0.738 1 0415
. H m + 0.0575 H 2 m 2
Kc (89 ) = K H
K m (1 + K H K m )
( 87)
] , T
( 89)
( 87)
(1 + K H K m )
H K m 0.0609 1 + 197
. H m 0.323 H 2 m 2
=
( 88)
] , T
Kc
Kc
( 88)
(1 + KH K m )
H K m 0.0078 1 + 8.38 H m 197

. H 2 m2
Kc
( 89 )
(1 + K H K m )
H K m 0.0124 1 + 4.05 H m 0.63 H 2 m 2
2
m
+ 0117
. m
Tm1
Tm1
)
)

( 90)
115
. ( 0.564) H m 1 0.959 + 0.773 2
K c (1 + K H K m )
Kc ( 90) = K H
, Ti (90) =
K m (1 + K H K m )
H K m 0.0335 1 + 0947
.
ln[ H m ] 1 + 19
. 1.07 2
107
. 1 0.466 H m + 0.0667 H 2 m 2
Kc ( 91) = KH
K m (1 + K H K m )
0.789 1 0.527 H m + 011

. H m
Kc ( 92) = K H
K m (1 + KH K m )
] , T
0.76 1 0.426 H m + 0.0551 H 2 m 2

Kc ( 93) = KH
K m (1 + K H K m )
( 92)
] , T
( 91)
] , T
)(
( 93)
Kc
( 91)
(1 + KH K m )
H K m 0.0328 1 + 2.21 H m 0.338 H 2 m 2

Kc
( 92)
(1 + K H K m )
H K m 0009
.
1 + 9.7 H m 2.4 H2 m 2
Kc
( 93)
(1 + KH K m )
H K m 0.0153 1 + 4.37 H m 0.743 H 2 m 2
2
m
+ 0117
. m
Tm1
Tm1
)(
0005
.
Tm1

( 86)
128
. ( 0.542) H m 1 0986
.
+ 0558
.
2
K c (1 + K H K m )
Kc (86) = KH
, Ti (86) =
K m (1 + K H Km )
H K m 0.0476 1 + 0.996 ln[ H m ] 1 + 2.13 113
. 2
114
. 1 0466
. H m + 0.0647 H 2 m 2
Kc (87 ) = K H
K m (1 + K H K m )
> 0649
.
+ 058
.
0471
. Ku
K m u
Ti (93)
> 0649
.
+ 058
.
0005
.
Tm1
Kc
Rule
Servo nearly
minimum IAE, ISE,
ITAE Hwang [60]
(continued)
Kc
( 94) 7
Kc
Minimum IAE
regulator ultimate
cycle - Shinskey [16]
page 151.
Model: Method 10.
Direct synthesis
Gain and phase
margin Hang et al.
[35]
Model: Method 4
Gain and phase
margin Ho et al.
[140]
Ti
Ti
( 95)
Ti
Gain and phase

margin Ho et al.
[142]
Model: Method 1
Comment
0471
. Ku
K m u
( 94 )
0649
.
+ 058
.
> 0613
.
+ 0613
.
0471
. Ku
K m u
(95)
2
m
0005
. m
Tm1
Tm1
0649
.
+ 058
.

m
+ 0117
. m
Tm1
Tm1
2
m
0005
. m
Tm1
Tm1
> 0613
.
+ 0613
.
0.38Tu
015
. Tu
Tm 2 Tm2 + m = 0.25
0.6766Ku
0.33Tu
019T
. u
Tm 2 Tm2 + m = 0.5
0.7874K u
0. 26Tu
0.21Tu
Tm 2 Tm2 + m = 0.75
Tm1
,
Am Km m
2Tm1 ,
0.5Tm1
Sample A m , m
A m = 1.5, m = 30 o
A m = 3.0 , m = 60 o
A m = 2 .0, m = 45o
A m = 4.0 , m = 67 .5o
p Tm1
A m Km
4 p m
2
2 p
A m = 2 .0, m = 45
A m = 5.0, m = 72 o
provided. Model has a

repeated pole ( Tm1 )
Tm 2
Tm1 > Tm 2 . Sample

A m , m provided
A m m + 05
. A m (A m 1)
1
Tm1
A m = 3.0 , m = 60 o
A m = 4.0, m = 70 o
Kc( 96)
Ti (96 )
Td ( 96)
A m = 2 .0, m = 45o
A m = 3.0 , m = 60 o
A m = 4.0, m = 70 o
Kc (97)
Ti (97 )
Td ( 97)
A m = 2 .0, m = 45o
A m = 3.0 , m = 60 o
A m = 4.0, m = 60 o
111
. 1 0467
. H m + 0.0657 H 2 m 2
Kc ( 95) = K H
K m (1 + KH K m )
p =
(A
2
m
1 m
m m
,
1
Tm Tm
Sample A m , m
provided *
n m < 2 m ; Sample
A m , m provided
2 m
] , T
( 97)
Td ( 97)
)(
( 95)
Kc
( 95)
(1 + KH K m )
H K m 0.0477 1 + 2.07 H m 0.333 H 2 m 2
m
( 97) 2
2 m
+ 2 p m , Ti
= Tm1
+ 2 p m ,
p Tm1
Tm1 p
Tm1
=

2 m + Tm1 2Tm1 p m
p
2

Tm1
( 96)
( 96)
m + 2 m , Ti = Tm1( m + 2 m ) , Td =
A m K m
Tm1
2( mTm1 + Tm1 2 m )
2 pTm1
Am
Kc
2
m
+ 0117
. m
Tm1
Tm1

( 94)
122
. ( 0.55) H m 1 0.978 + 0.659 2
K c (1 + K H K m )
Kc ( 94) = K H
, Ti (94) =
K m (1 + K H K m )
H K m 0.0421 1 + 0969
.
ln[ H m ] 1 + 2.02 111
. 2
Kc ( 96) =
,
2
0.6173Ku
Model: Method 1
Gain and phase
margin Ho et al.
[141]
Model: Method 1
Td
8m m
+ 2 +
A m 2 1 2A m ( A m 1)
T
m1
* m <
4A m
Kc
Rule
Ti
Td
Comment
m > 0.7071 or
m Tm1
Km m
Wang et al. [143]
0.05 <
05
. Tm1
m
2 m Tm1
07071
.
m
Tm1 2 m 2 1
0.7071m
Tm1 2 m 2 1
Model: Method 8
or 0.05 <
m m
Tm1
Minimum of
2
2 Tm1
e
Km
Gain and phase

margin Leva et al.
[34]
Tm m
m
Kc
07358
.
m Tm1
Km m
( 98) 8
Ti
( 98 )
0.25Ti
> 1, m 1
m m
Tm1
< 015
. ,
> 1, m < 1
m 0.7071 and
05
. Tm1
m
2 m Tm1
< 015
.
0.15
m m
Tm1
m 1 or
m > 1 with
05
. m >
3 m
2
m + m 1
Tm1
( 98)
Model: Method 5
m > 1 with
Kc( 99)
Ti (99 )
3 m
2
m m 1
Tm1
< 0.5 m
3 m
2
m + m 1 ,
Tm1
Td ( 99)
m = 70 0 at least
Kc
( 98)
Ti( 98)
z=
cn Ti
Km
( 1 + T ) + 4
(1 T T ) + T
2
cn
m1
i d
cn
cn
Tm1
2
, cn =
cn
2 T ( 05
1
m m m1 . m )
4.07 m + tan 1
2
2
2
m
( 0.5 m ) Tm1 m

2
2 m cnTm1
2
1
, Kc ( 99) = cnTm(99
=
tan 05
. cn m + m 05
. tan 1
2
2
cn

cn Tm1 1
K mTd )
cn
cn Tm1
1
tan m 05
. + cn m + tan
m + m 2 1

, Ti (99 ) =
Tm1z + m + m 2 1
z m + m 2 1
cn 2 +
1
2 m m 2 1 1
Tm12
,
2
2
cn + z
, Td (99) =
Tm1
Tm1z + m + m2 1
Rule
Gain and phase
margin Wang and
Shao [144]
Kc
other tuning rules

obtained using
authors method
Td
m Tm1
Km m
2 m Tm1
05
.
Tm1
m
Am = 3 , m = 60 0
2.094
mTm1
K m m
2 m Tm1
05
.
Tm1
m
A m = 15
. , m = 30 0
1571
.
mTm1
K m m
2 m Tm1
05
.
Tm1
m
A m = 2 , m = 450
0.785
m Tm1
K m m
2 m Tm1
05
.
Tm1
m
A m = 4 , m = 67.50
0.628
mTm1
K m m
2 m Tm1
05
.
Tm1
m
A m = 5 , m = 72 0
Pemberton [145]
Model: Method 1
2( Tm1 + Tm2 )
3Km m
Tm1 + Tm2
Tm1Tm2
Tm1 + Tm 2
Pemberton [24]
(Tm1 + Tm2 )
K m m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2
2( Tm1 + Tm 2 )
3Km m
Tm1 + Tm2
Model: Method 1
Pemberton [145]
Model: Method 1
Comment
1047
.
Model: Method 8
Authors quote
tuning rule for
Am = 3 , m = 60 0 ;
Ti
Tm1 + Tm 2
4
01
.
m
10
. ;
Tm1
0.2
m
10
.
Tm 2
01
.
m
10
.
Tm1
0.2
m
10
.
Tm 2
Suyama [100]
Model: Method 1
Tm1 + Tm 2
2K m m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2
Smith et al. [146]

Model: Method not
stated
Chiu et al. [29]
Tm1 + Tm 2
K m ( + m )
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2
Tm1 + Tm2
K m (1 + m )
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2
Tm 1
K m (1 + m )
variable; suggested
values are 0.2, 0.4, 0.6
and 1.0.
Tm1 > Tm2 . = pole of
Tm1
Tm2
2 m Tm1
K m (1 + m )
2 m Tm1
Tm1
2 m
specified closed loop

overdamped response.
Underdamped
response
Gorez and Klan

[147a]
Model: Ideal
Miluse et al. [27a]
2 m Tm1
K m (2 m Tm1 + m )
2 m Tm1
Tm1
2 m
Non-dominant time
delay
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2
Closed loop overshoot

= 0%
Model: Method not

specified
0. 368 (Tm1 + Tm 2 )
K m m
0. 514 (Tm1 + Tm 2 )
K m m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2

= 5%
0. 581( Tm1 + Tm 2 )
Km m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2

= 10%
0. 641( Tm1 + Tm 2 )
Km m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2

= 15%
Model: Method 6
Wang and Clemens
[147]
Model: Method 9
Overdamped
process;
Tm1 > Tm 2
Rule
Kc
Ti
Td
Comment
Miluse et al. [27a]

(continued)
0. 696 (Tm1 + Tm 2 )
K m m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2

= 20%
0. 748 (Tm1 + Tm 2 )
K m m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2

= 25%
0. 801( Tm1 + Tm 2 )
Km m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2

= 30%
0. 853( Tm1 + Tm2 )

K m m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2

= 35%
0. 906 (Tm1 + Tm 2 )
K m m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2

= 40%
0. 957 (Tm1 + Tm 2 )
K m m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2

= 45%
1.008 ( Tm1 + Tm 2 )
K m m
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2

= 50%
2 m Tm1
Tm1
2 m

= 0%
Model: Method not

specified
0. 736 mTm1
Km m
1.028 m Tm1
K m m
2 m Tm1
Tm1
2 m

= 5%
Underdamped
process;
0.5 < m 1
1.162 m Tm1
K m m
2 m Tm1
Tm1
2 m

= 10%
1.282 m Tm1
K m m
2 m Tm1
Tm1
2 m

= 15%
1.392 m Tm1
K m m
2 m Tm1
Tm1
2 m

= 20%
1.496 m Tm1
K m m
2 m Tm1
Tm1
2 m

= 25%
1.602 m Tm1
K m m
2 m Tm1
Tm1
2 m

= 30%
1.706 m Tm1
K m m
2 m Tm1
Tm1
2 m

= 35%
1.812 m Tm1
K m m
2 m Tm1
Tm1
2 m

= 40%
1.914 m Tm1
K m m
2 m Tm1
Tm1
2 m

= 45%
2.016 m Tm1
K m m
2 m Tm1
Tm1
2 m

= 50%
0. 736 Tm1
K m m
2Tm1
0.5Tm1
Tm1 = Tm 2
Miluse et al. [27a]
Miluse et al. [27b]

Model: Method 14
Re-tuning rule. 0 =
2
Seki et al. [147b]
c
tan c + 2 + tan 2 c
cos
Model: Method not G c ( j 0 )

c
2
0
c
specified
0.5Ti
freq. when controlled

system goes unstable
(crossover freq.);
c = new crossover
freq. ; c = phase lag
Rule
Autotuning
Landau and Voda
[148]
Model: Method not
relevant
Kc
Ti
Td
Comment
4+
4 2 2 G p ( j )
135
4+ 1
135
4
1
4 + 135
1 2 ;
m < 0.25Tm1
3.2
u
08
.
u
u m 016
.
4
u
1
u
016
. < u m 0.2
Tm 1 + Tm2 + 05
. m
Tm1 Tm2 + 0.5( Tm1 + Tm2 ) m
varies graphically
with m (Tm1 + Tm 2 )
G p ( j u )
19
.
G p ( j u )
Robust
Brambilla et al. [48]
values deduced from
graph
Model: Method 1
Chen et al. [53]
Tm1 + Tm 2 + 0.5 m
K m m ( 2 + 1)
m (Tm1 + Tm 2 )
0.1
0.50
0.2
0.47
0.5
0.39
1.00 m Tm
mK m
Tm1 + Tm 2 + 05
. m
0.1 m ( Tm1 + Tm 2 ) 10
m (Tm1 + Tm 2 )
m (Tm1 + Tm 2 )
1.0
2.0
5.0
0.29
0.22
0.16
10.0
0.14
A m = 3.14 ,
2 m m
m
2 m
1.22 m Tm
mK m
2 m m
m
2 m
A m = 2.58 ,
1.34 m Tm
mK m
2 m m
m
2 m
A m = 2. 34 ,
1.40 m Tm
mK m
2 m m
m
2 m
A m = 2. 24 ,
1.44 m Tm
mK m
2 m m
m
2 m
A m = 2.18 ,
1.52 m Tm
mK m
2 m m
m
2 m
1.60 m Tm
mK m
2 m m
m
2 m
m = 61. 40 ,
Ms = 1. 00
Model: Method 1
m = 55. 00 ,
Ms = 1. 10
m = 51.6 0 ,
Ms = 1. 20
m = 50.0 0 ,
M s = 1. 26
m = 48. 70 ,
Ms = 1. 30
A m = 2. 07 ,
m = 46. 50 ,
Ms = 1. 40
A m = 1.96 ,
m = 44.10 ,
Ms = 1. 50
2
Lee et al. [55]

Model: Method 1
Ti
K m ( 2 + m )
2 mTm1
2 m
2( 2 + m )
2
Ti 2 mTm1 +
Tm1
Ti
m
6Ti ( 2 + m )
3
Desired closed loop

e m s
response =
( s + 1) 2
Rule
Kc
Lee et al. [55]

(continued)
Ti
K m ( 2 + m )
Ti
K m( + m )
Ti
K m( + m )
Ti
2 2 m 2
2( 2 + m )
Tm1 + Tm2
2 mTm1
Td
m
+
2( + m )
Tm1 + Tm2
m
+
2 ( + m )
2
Comment
Ti Tm1 Tm2 +
Tm1Tm2
Ti
Desired closed loop

e m s
response =
( s + 1) 2
m
6Ti ( 2 + m )
3
m3
Tm1 2
6( + m )
Ti 2 m Tm1
Ti
Desired closed loop

e m s
response =
( s + 1)
m3
Tm1Tm 2
6 + m
Ti ( Tm1 + Tm2 )
Ti
Desired closed loop

e m s
response =
( s + 1)
Km e s
K m e s
or
- filtered controller
(1 + sTm1 )(1 + sTm 2 ) Tm12 s2 + 2 m Tm1s + 1
m
1
1
G c ( s) = Kc 1 +
+ Td s
. 1 tuning rule.
Ti s
Tf s + 1
Rule
Kc
Ti
2Tm1 + m
2( + m ) K m
Tm1 + 05
. m
Td
Comment
Robust
Hang et al. [35]
Model: Method 4
Tm1m
2Tm1 + m
Model has a repeated

pole (Tm1 ) .
> 025
. m .
Tf =
m
2( + m )
Km e s
K m e s
or
- filtered controller
(1 + sTm1 )(1 + sTm 2 ) Tm12 s2 + 2 m Tm1s + 1
m
b s+1
1
G c ( s) = Kc 1 +
+ Td s 1
. 1 tuning rule.
Ti s
a 1s + 1
Rule
Robust
Jahanmiri and Fallahi
[149]
Model: Method 6
Kc
Ti
Td
2 m Tm1
K m ( m + )
= 0.25 m + 01
. m Tm1
2 m Tm1
Tm1
2 m
Comment
b1 = 0.5 m
a1 =
m
2( m + )
Km e s
K m e s
or
- classical controller
(1 + sTm1 )(1 + sTm 2 ) Tm12 s2 + 2 m Tm1s + 1
m
1 1 + Td s
. 7 tuning rules.
G c ( s) = Kc 1 +
Ti s 1 + Td s
N
Kc
Rule
Regulator tuning
Ti
Td
1.5
m 15
. e
m1

159
Kc
( 100 )
0800
. Tm1
Km m
0770
. Tm1
K m m
0833
. Tm1
Km m
** Minimum IAE
Shinskey [17]
059
. Ku
085
. Tm1
K m m
087
. Tm1
Km m
100
. Tm1
K m m
125
. Tm1
K m m
Kc (100) =
48 + 57 1 e
m2
25
. Tm1
K m m
** Minimum IAE Shinskey [17]
1 + 0.9 1 e
Model: Open loop

** Minimum IAE Shinskey [59]
Comment
1.2 Tm1
1.2 T
0.56 m 1 e
m1
Tm2
3
m
0.6Tm2
m + 0.2Tm2
m + 0.2Tm2
15
. ( Tm2 + m )
060
. ( Tm2 + m )
12
. ( Tm2 + m )
0.70( Tm 2 + m )
075
. ( Tm2 + m )
060
. ( Tm2 + m )
0. 36Tu
0. 26Tu
198
. m
086
. m
2.30 m
165
. m
2.50m
200
. m
2.75m
2.75 m
100
2
Tm 2
Km m
Tm2
1 + 0.34
0.2

Tm1
m
m
Tm2
>3
m
Tm 2
= 025
.
Tm 2 + m
Tm 2
= 0.5
Tm2 + m
Tm 2
= 0.75
Tm2 + m
m
T
= 0.2 , m 2 = 0.2
Tm1
Tm1
m
T
= 0.2 , m 2 = 01
.
Tm1
Tm1
Tm 2
= 0.2
Tm1
Tm 2
= 0.5
Tm1
Tm 2
= 10
.
Tm1
Rule
Kc
Ti
Td
Minimum ISE
McAvoy and
Johnson [83]
deduced from graph
Km
Model: Method 1
m
Tm1
m
Tm1
Comment
m
Tm1
N = 20
1
1
1
0.5 0.7 0.97 0.75

4.0 7.6 3.33 2.03
10.0 34.3 5.00 2.7
4
4
0.5
4.0
3.0 1.16 0.85

22.7 1.89 1.28
7
7
0.5
4.0
5.4 1.19 0.85

40.0 1.64 1.14
N = 10
1
1
1
0.5 0.9 1.10 0.64

4.0 8.0 4.00 1.83
10.0 33.5 6.25 2.43
4
4
0.5
4.0
3.2 1.33 0.78

23.9 2.17 1.17
7
7
0.5
4.0
5.9 1.39 1.04

42.9 1.89 1.37
Direct synthesis
3
Astrom et al. [30]
Method 13
Smith et al. [26]

deduced from graph
[ = m ( Tm1 + Tm2 ) ]
Model: Method not
stated
3 m
K m 1 +
Tm1 + Tm 2
Tm1 + Tm2
Tm 1Tm2
Tm1 + Tm 2
Tm1
Tm2
N = 8 - Honeywell
UDC6000 controller
Tm1
K m m
6
1.75
1.75
1.5
1.0
0.02
0.27
0.07
0.11
0.25
0.51
0.46
0.36
0.33
0.46
3
1.75
1.5
1.5
1.0
004
.
0.13
0.33
0.08
0.17
0.50
0.42
0.44
0.28
0.48
2
1.75
1.5
1.0
1.0
006
.
0.07
0.17
0.50
0.13
0.45
0.39
0.38
0.40
0.49
N = 10
Km e s
K m e s
or
- alternative classical
(1 + sTm1 )(1 + sTm 2 ) Tm12 s2 + 2 m Tm1s + 1
m
1 1 + NTd s
controller G c ( s) = Kc 1 +
. 1 tuning rule.
Ti s 1 + Td s
Rule
Kc
Ti
Td
Comment
******
0.7
Hougen [85]
Model: Method 1
08
. Tm1 Tm2
m
0.8
0. 3
084
. Tm1 Tm 2
m
05
. Tm1 + Tm 2
m Tm1 Tm 2
N=10
0.2
0.53Tm1 + 1.3Tm2
0.08( m Tm1Tm2 )
0.28
N=30
Km e s
K m e s
or
- alternative non(1 + sTm1 )(1 + sTm 2 ) Tm12 s2 + 2 m Tm1s + 1
m
1
interacting controller 1: U( s) = K c 1 +
E ( s) Kc Td sY ( s) . 3 tuning rules.
Ti s
Kc
Rule
Regulator tuning
159
Model: open loop

119. Model: method
not specified.
Kc (101) =
38 + 401 e
Td
Comment

1
Kc
( 101)
m1
m 0.5 + 1. 4[ 1 e 1.5 m ]
m2
1 + 0.48 1 e m
1.2 Tm1
0.42 m 1 e m +
0.6Tm2
Tm2
3
m
m + 0.2Tm2
m + 0.2Tm2
118
. Tm1
Km m
2.20 m
0.72 m
m
T
= 0.2 , m 2 = 01
.
Tm1
Tm1
125
. Tm1
K m m
2.20m
110
. m
Tm 2
= 0.2
Tm1
167
. Tm1
K m m
2.40 m
165
. m
Tm 2
= 0.5
Tm1
25
. Tm1
K m m
2.15 m
2.15 m
Tm 2
= 10
.
Tm1
085
. Ku
035
. Tu
017
. Tu
m
T
= 0.2 , m 2 = 0.2
Tm1
Tm1
333
. Tm 1
Km m

121. Model: method
not specified.
Ti
1.5Tm1
100
2
Tm 2
Km m
Tm2
1 + 0.34
0.2

Tm1
m
m
Tm 2
>3
m
Km e s
K m e s
or
- series controller
(1 + sTm1 )(1 + sTm 2 ) Tm12 s2 + 2 m Tm1s + 1
m
1
G c ( s) = K c 1 +
(1 + Td s) . 1 tuning rule.
Ti s
Rule
Kc
Regulator tuning
Least mean square
error - Haalman [23]
Model: Method 1
Ti
Td
Comment
2Tm1
3 m K m
Tm1
Tm2
Tm1 > Tm2 . Maximum

sensitivity = 1.9, Gain
margin = 2.36, Phase
margin = 50 0
Km e s
K m e s
or
- non-interacting
(1 + sTm1 )(1 + sTm 2 ) Tm12 s2 + 2 m Tm1s + 1
m
1
Td s
. 2 tuning rules.
E( s)
controller U (s) = K c 1 +
Y
(
s
)
T
s
T
s
d
i
1+
Kc
Ti
Td
Rule
Servo Min. IAE Huang et al. [18]
Model: Method 1
Kc (102 ) =
Kc
( 102 ) 2
Ti
( 102)
Td
Comment
0<
( 102 )
0.1
Tm 2
1;
Tm1
m
Tm1
1 ; N =10
0.0865
m
1
T
T
7.0636 + 66.6512 m 137 .8937 m2 122 .7832 m m2 2 + 261928
.
Km
Tm1
Tm1
Tm1
Tm1
2.6405
1.0309
2.345
1.0570
m
Tm 2
T
T
T
1
336578
.
+ 30098
.
10.9347 m 2
+ 141511
. m2
+ 29.4068 m2
Km
Tm1
Tm1
Tm1
Tm1
m
0.9450
0.9282
0.8866
Tm2 m
Tm 2 m
m Tm2
1
34.3156
+
701035
.
+ 152.6392
Km
Tm 1 Tm1
Tm1 Tm1
Tm1 Tm1
1
+
Km
Ti
( 102)
0 .8148
T
T
47.9791 m m 2
57.9370e + 10.4002e T
Tm1 Tm1
m2
m Tm 2
m1
m1
Tm1 2
+ 6.7646e

+ 7.3453 m
Tm
0.4062
2
2
3
m
Tm2
m
m
Tm2
m Tm2
= Tm1 0.9923 + 0.2819
0.2679
1.4510
+ 0.6424
0.6712
+ 2 .504

2
Tm1
Tm1
Tm1
Tm1
Tm1
Tm1
2
2
3
4
4
T
T
T
m
m Tm2
+ Tm1 2.5324 m 2 m + 2.3641 m m 2 + 2.0500 m 2 18759
.
+
08519
.

Tm1 Tm1
Tm1 Tm1
Tm1
Tm1
Tm1 Tm1
3
2
2
3
4
m Tm2
Tm 2
Tm2 m
T
+ Tm 1 13496
.
.
.
34972

2.4216 m m 2 31142

Tm1 Tm1
Tm1 Tm1
Tm1 Tm1
Tm1
5
4
2
3
2
3
5
m
Tm2 m
m Tm 2
Tm 2
Tm2 m
+ Tm 1 05862
.
.
.
+ 0.0797
+ 0.985

+ 12892

+ 12108

Tm1 Tm1
Tm1
Tm1 Tm1
Tm1 Tm1
Tm1
Td
( 102)
2
2
3
m
Tm2
m
m
Tm 2
m Tm 2
= Tm1 0.0075 + 0.3449

+ 0.3924
0.0793
+ 0.6485
+ 2.7495
+ 0.8089

2
Tm1
Tm1
Tm1
Tm1
Tm1
Tm1
+ Tm1 9.7483 m 2 m + 3.4679 m

Tm1 Tm1
Tm1
Tm2
Tm2
m
.
.
58194
10884
Tm1
Tm1
Tm1
3
2
2
3
4
m Tm 2
Tm 2
T
T
+ Tm 1 12.0049 m 2 m 14056
.
.

3.7055 m m 2 + 100045

Tm1 Tm1
Tm1 Tm1
Tm1 Tm1
Tm1
5
4
2
3
2
3
m
Tm2 m
m Tm 2
Tm 2 m
+ Tm 1 0.3520
.
6.3603
32980

+ 7.0404

Tm1 Tm1
Tm1
Tm1 Tm1
Tm1 Tm1
4
5
6
m Tm 2
Tm 2
m
Tm 2
+ Tm1 14294
.
69064
.
+
0
.
0471
.
+ 11839
T
T
T
T
Tm1
m1 m1
m1
m1
m
Tm 2
.
+ 17087
T
m1
Tm1
Kc
Rule
Servo-Min. IAE Huang et al. [18]
Kc
( 103) 3
Ti
(N=10)
Ti
Td
( 103)
( 103)
Td
Comment
0 .4 m 1 ; 0 .05
4
2
3
3
2
4
5
m Tm 2
m Tm2
m Tm 2
m Tm 2
+ Tm1 1. 7444
.
.
.

12817

21281

+ 15121

Tm1 Tm1
Tm1 Tm1
Tm1 Tm1
Tm1 Tm1
Kc (103) =
1.9009
81727
.
32.9042 m + 319179
.
m + 38.3405 m m + 0.2079 m
Km
Tm1
Tm1
Tm1
0.1571
1.2234

m
1
29.3215 m
+ 359456
.
214045
.
m0 .1311 + 51159
.
m1.9814 219381 m1.737
Km
Tm1
Tm 1
0.1303
1.2008
m
m
1
17.7448 m
+
+ 268655
.
m
52.9156 m m1.1207
Km
Tm1
Tm1
Tm1
1
+
Km
Ti
( 103)
T
22.4297 m m 0.3626 33331
.
e
Tm1
m
m1
+ 85175
.
e
15312
.
e
mm
Tm1
+ 08906
.
m Tm1
2
2
2
= Tm1 11731
.
+ 6.3082 m 0.6937 m + 8.5271 m 24 .7291 m m 6.7123 m m + 7 .9559 m
Tm1
Tm1
Tm1
Tm1
3
4
3
m
m
m

+ Tm1 32.3937
.
.
m2 m
271372
+ 166.9272 m
+ 363954
Tm1
Tm1
Tm1
Tm1
m
2 m
3
+ Tm1 94.8879 m
.
m 3 + 29.9159 m 4
22.6065 m 16084
Tm1
Tm1
5
4
3
2

m
3
+ Tm 1 49.6314 m 84.3776 m m 938912
.
m 2 m + 1101706
.
Tm1
Tm1
Tm1
Tm1
6
5
m 4

5
+ Tm 1 251896
.
.
m m + 55268
.
m6
m 19.7569 m 12.4348 m 117589
Tm1
Tm1
Tm1
4
3
2
m
m
m
m
2
3
4
+ Tm 1 68.3097
.
.
m 5
m 17.8663
m 225926
m + 95061
Tm1
Tm1
Tm1
Tm1
Td
( 104)
2
3
m
m
m
m
2
= Tm1 0.0904 + 0.8637

0.1301 m + 4 .9601
+ 0.7170 m 12.5311
+ 14 .3899 m

Tm1
Tm1
Tm1
Tm1
2
4

+ Tm1 42.5012 m m 214907
.
m2 m 69555
.
m 3 12.3016 m
Tm1
Tm1
Tm1
3
2
+ Tm1 102.9447 m m + 7.5855 m 2 m + 19.1257 m m 3 + 17.0952 m 4

Tm1
Tm1
Tm1
5
4
3
2
m
m
m
2 m
3
+ Tm1 108688
.
.
m
17.2130 m
1100342
+ 50.6455
m
Tm1
Tm1
Tm1
Tm1
6
5

+ Tm 1 16.7073 m m 4 16.2013 m5 0. 0979 m 109260
.
m m + 54409
.
m6
Tm1
Tm1
Tm1
m
1
Tm1
Rule
Kc
Ti
Td
Regulator Min. IAE

- Huang et al. [18]
Model: Method 1
Kc(104 ) 4
Ti (104)
Td (104 )
Comment
Tm 2
1;
Tm1
0<
m
0.1
Tm1
1 ; N =10
4
3
2
m
m
m
m 5
2
3
4
+ Tm 1 29.4445
.
.
.
m
m + 216061
m 241917
m + 62798
Tm1
Tm1
Tm1
Tm1
Kc (104) =
0.8058
m
1
m
Tm2
m Tm2
01098
.
86290
.
+ 766760
.
33397
.
+
11863
.
2
Km
Tm1
Tm1
Tm1
Tm1
0.6642
2 .1482
0.8405
2.1123
m
m
T
Tm2
1
231098
.
+ 203519
.
52.0778 m 2
121033
.
Km
Tm1
Tm1
Tm1
Tm1
0.5306
1.0781
0.4500
1
T
T
T
9.4709 m m2
+ 13.6581 m 2 m
19.4944 m2 m
Km
Tm1 Tm1
Tm1 Tm1
Tm1 Tm1
1.1427
T
1
T
28.2766 m m2
+
191463
.
e T + 8.8420e T
Km
Tm1 Tm1
Ti
( 104)
m2
m1
m1
m Tm 2
Tm1 2
+ 7.4298e
114753
.
Tm 2
2
2
3
T
T
= Tm1 0.0145 + 2 .0555 m + 0.7435 m2 4.4805 m + 1.2069 m m2 2 + 0.2584 m 2 + 7.7916 m
Tm1
Tm1
Tm1
Tm1
Tm1
Tm1
2
2
3
4
3
Tm 2
m
Tm 2 m
m Tm2
Tm 2 m
+ Tm1 6.0330
.
+ 3.9585
30626
7.0263
+ 7.0004

Tm1 Tm1
Tm1 Tm1
Tm1 Tm1
Tm1
Tm1
2
2
3
4
5
T
m Tm 2
Tm 2
m
+ Tm1 2.7755 m 2 m 15769
.
+
31663
.
+
2
.
4311
T m1 Tm1
Tm1 Tm1
Tm1
Tm1
T
T
m Tm2
T
+ Tm 1 0.9439 m 2 2.4506 m 2 m 0.2227 m2 m + 19228

.

0.5494 m
Tm1 Tm 1
Tm1
Tm1
Tm1 Tm1
Tm1 Tm1
5
Td
( 104)
2
2
3
m
Tm2
m
m
Tm 2
mTm 2
= Tm1 0.0206 + 0.9385
+ 0.7759
2.3820
3.2336
+ 2 .9230
+ 7.2774

2
Tm1
Tm1
Tm1
Tm1
Tm1
Tm1
T
+ Tm 1 9.9017 m 2
Tm1
+ 2.7095 m
T
T
m1
m1
Tm 2
Tm2
m
.
.
+ 61539
111018
T
T
m1
m1
Tm1
3
2
2
m Tm2
T
+ Tm 1 10.6303 m2 m + 57105
.

7.9490 m
Tm1 Tm1
Tm1
Tm1 Tm1
Tm 2
T
6.6597 m 2
Tm1
Tm1
5
4
2
3
2
3
m
T
m Tm2
T
+ Tm 1 80849
.
.
4.4897 m 2 m 7.6469 m2 m + 21155

Tm1 Tm1
Tm1
Tm1 Tm1
Tm1 Tm 1
4
5
6
5
6
m Tm 2
Tm 2
m
Tm 2
Tm 2 m
+ Tm 1 50694
.
.
.
+ 4.1225
2.274
+ 0519
11295

Tm1 Tm1
Tm1 Tm1
Tm1
Tm1
Tm1
4
2
3
3
2
4
T
T
m Tm 2
m
+ Tm 1 2.2875 m m2 + 0.9524 m m 2 16307
.
.

09321
T
T
T
T
T
T
T
m1 m1
m1 m1
m1 m1
m1
Tm2
Tm1
Tm2
Tm1
Rule
Regulator - Min. IAE Huang et al. [18]
Model: Method 1
Kc
( 105)
Kc
Kc
( 105) 5
Ti
Ti
Td
( 105)
Td
( 105)
Comment
[N=10]
0 .4 m 1 ; 0 .05
m
1
Tm1
3
0.086
m
m
m
1
m
=
357307
.
1419
.
+ 14023
.
m + 6.8618 m
.
m
0.9773
+ 555898
Km
Tm1
Tm1
Tm1
Tm
2
1. 6624
0.6951

m

1
m
33093
.
m m + 538651
.
m 2 + 114911
.
m3 + 08778
.
29.8822 m
Km
Tm 1
Tm1
Tm1
Tm1
0.4762
2.1208

1
53535
. m
16.9807 m1.1197 254293
.
m1.4622 01671
.
m 58981 + 0.0034 m m
Km
Tm1
Tm

1
m
m 1.2103
m Tm1
3.0836
T
250355
+
.
m
54.9617
m
01398
.
e
82721
.
e + 63542
.
e T + 10479
.
Km
Tm1
Tm1
m
m1
Ti
( 105)
m1
2
3
m
m
m
m
2
= Tm1 0.2563 + 11.8737

1.6547 m 16.1913
+ 3.5927 m + 19.5201
9 .7061 m

Tm1
Tm1
Tm1
Tm1
2
4

+ Tm 1 14.5581 m m + 2.939 m 2 m 0.4592 m 3 34.6273 m
Tm1
Tm1
Tm1
3
2
2 m
3
4
+ Tm 1 50.5163 m
+ 8.9259 m
+ 8.6966 m m 6.9436 m
Tm1
Tm1
Tm 1
5
4
3
2
m
m
m
2 m
3
+ Tm 1 27.2386
.
20.0697 m
42.2833 m
+ 85019
Tm1
Tm1
Tm1
Tm1
6
5

+ Tm 1 12.2957 m m 4 + 8.0694 m 5 7.7887 m + 2.3012 m m 2.7691 m 6
Tm1
Tm1
Tm1
4
3
2
m
m
m
2
3
4
5
+ Tm 1 88984
.
.
.
m + 102494
m 54906
m + 4.6594 m m
Tm1
Tm1
Tm1
Tm1
Td
( 105)
2
3
m
m
m
m
2
= Tm1 0.021 + 3.3385
+ 0.185 m 0.5164
0.8815 m + 0.584
0.9643 m

Tm1
Tm1
Tm1
Tm1
2
4

m
2
3
+ Tm1 12513
. m m + 13468
.
m m + 2.3181 m 52368
.
Tm1
Tm1
Tm1
3
2
2
3
4
+ Tm1 153014
.
m m + 119607
.
m m 2.0411 m m 31988
.
m
Tm1
Tm1
Tm1
5
4
3
2
m
m
m
2 m
3
+ Tm1 34675
.
0.8219 m
15.0718m
18859
.
Tm1
Tm1
Tm1
Tm1
6
5
4

5
6
+ Tm1 0.4841 m m + 2.2821 m 0.9315 m + 0529
. m m 06772
.
m
Tm1
Tm1
Tm1
4
3
2
m
m

2
3
4
5
+ Tm1 14212
.
m + 71176
.
m 2.3636 m m + 0.5497 m m
Tm1
Tm1
Tm1
Tm1
Km e s
K m e s
or
- ideal controller with
(1 + sTm1 )(1 + sTm 2 ) Tm12 s2 + 2 m Tm1s + 1
m
set-point weighting U(s) = Kc Fp R ( s) Y( s) +
Kc
Rule
Kc
[ Fi R(s) Y(s)] + Kc Td s[ Fd R (s) Y(s)] . 1 tuning rule.
Ti s
Ti
Td
Comment
05
. Tu
Fi = 1
0125
. Tu
Repeated pole
01
. < m < 3;
Tm
Ultimate cycle
06
. Ku
Oubrahim and
Leonard [138]
Fp = 13
.
Model: Method not

specified
16 Km Ku
17 + Km Ku
Fd = Fp
06
. Ku
Fp =
38
29 + 35Km K u
05
. Tu
Fi = 1
0125
. Tu
Fd = Fp
10% overshoot
Repeated pole
01
. < m < 3;
Tm
20% overshoot
Km e s
K m e s
or
- non-interacting
(1 + sTm1 )(1 + sTm 2 ) Tm12 s2 + 2 m Tm1s + 1
m
1
controller U( s) = Kc b + [ R (s) Y(s) ] ( c + Tds)Y( s) . 1 tuning rule.
Ts
i
Rule
Kc
Ti
Td
Comment
Kc(105a) 6
154
. Ti ( 105a )
Td (105a)
c = K c ( 105a ) 1 K m ;
Direct synthesis
b = 0.198
Hansen [150]
zero overshoot
b = 0.289
Model: Method 1
127
. Ti (105a )
Kc(105a)
Td (105a)
c = K c ( 105a ) 1 K m ;
minimum IAE
b = 0.143
175
. Ti (105a )
Kc(105a)
Td (105a)
c = K c ( 105a ) 1 K m ;
conservative tuning
Kc
Td
(105a )
( 105a )
2 Tm1Tm 2 + (Tm1 + Tm2 ) m + 0.5 m2
[
2[ T
9 Km Tm1Tm2 m + 0.5(Tm1 + Tm2 ) m + 0167

. m
m1Tm2
+ (Tm1 + Tm 2 ) m + 05
. m2
3 Tm1Tm2 m + 05
. ( Tm1 + Tm 2 ) m 2 + 0167
. m3
3K m Tm1Tm 2 m + 0.5(Tm1 + Tm2 ) m + 0167

. m
2
3 2
, Ti (105a ) =
Tm1 + Tm2 m
Km
[T
m1Tm 2
+ (Tm1 + Tm2 ) m + 05
. m
],
Km e s
K m e s
or
- ideal controller with
(1 + sTm1 )(1 + sTm 2 ) Tm12 s2 + 2 m Tm1s + 1
m
1
Td s
filtered derivative G c ( s) = Kc 1 +
+
sT
Ts
i
1+ d
N
Rule
Kc
Ti
p Tm 1
A m Km
4 p 2 m
. 2 tuning rules.
Td
Comment
Tm2
provided. N = 20.
Tm1 > Tm2
Direct synthesis
Hang et al. [151]
2 p
Model: Method 1
A m = 2 .0 , m = 45o
A m = 3.0 , m = 60
Robust
Hang et al. [151]
Model: Method 1
Tm1
K m( + m)
Sample A m , m
1
+
Tm1
A m = 4.0 , m = 67.5 o
A m = 5.0, m = 72 o
Tm1
Tm2
p =
A m m + 0.5A m (A m 1)
(A
2
m
1 m
N = 20. Tm1 > Tm2
Table 80: PID tuning rules SOSPD model
K m e s
Tm1 s2 + 2 m Tm1s + 1
2
1
Td s
U (s) = K c 1 +
+
Tis
T
1+ d
Kc
Rule
T
s
d
E( s) K c +
Td
s
1+
s
Ti
Servo/regulator
tuning
Td
Taguchi and Araki

[61a]
Kc
(105 b) 7
Ti
(105 b )
Td
(105b )
(105 c )
Ti
(105 c )
Td
(105 c )
Kc

Pagola [134a]
0.7236 K u
0.5247 Tu
0.1650 Tu
Model: Method 15
Kc
Td
(105 b)
(105 b)
1
0. 6978
=
1.389 +
m
Kc
[
0.02295 ] 2
Tm
Td
(105c )
(105 c )
m
1.0 ; m = 1. 0
Tm
20% ; settling time
settling time of
et al. [10]
m
1.0 ; m = 0.5
Tm
20% ; settling time
settling time of
et al. [10]
= 0. 5840 , = 1 ,
N = 10,
c = 139.65 0
K m = 1 ; Tm = 1 ;
m =1
2
3
(105 b)
0.02453 + 4.104 m 3. 434 m + 1.231 m
,
T
=
T
i
m
Tm
Tm
Tm
2
3
4
m
m
m
m
= Tm 0.03459 + 1. 852
2.741 + 2.359 0. 7962 ,
Tm
Tm
Tm
T m
= 0. 6726 0.1285
Kc
Comment

0.1371 m + 0.07345 m , = 0. 8665 0.2679 m + 0.02724 m
Tm
T
T
T
m
m
m
Tm
1
0.5013
=
0.3363 +
m
Kc
[
0.01147 ] 2
Tm
2
3
(105 c )
0.02337 + 4.858 m 5.522 m + 2.054 m
,
T
=
T

m
i
Tm
Tm
Tm
2
3
m
m
m
= Tm 0.03392 + 2. 023
1.161 + 0.2826 ,
Tm
Tm
Tm
= 0. 6678 0. 05413

0. 5680 m + 0. 1699 m , = 0. 8646 0.1205 m 0.1212 m
Tm
Tm
Tm
Tm
Tm
Rule
Kc
Ti
Td

Pagola [134b]
0.803 K u
0.509 Tu
0.167 Tu
c = phase
K m = 1 ; Tm = 1 ;
0.1 < m < 10
0.727 K u
0.524 Tu
0.165 Tu
0.672 Ku
0.532 Tu
0.161Tu
0.669 Ku
0.486 Tu
0.170 Tu
Model: Method 15
0.600 Ku
0.498 Tu
0.157 Tu
0.578 K u
0.481Tu
0.154 Tu
0.557 K u
0.467 Tu
0.149 Tu
0.544 Ku
0.466 Tu
0.141Tu
0.537 K u
0.444 Tu
0.144 Tu
0.527 K u
0.450 Tu
0.131Tu
0.521K u
0.440 Tu
0.129 Tu
0.515 K u
0.429 Tu
0.126 Tu
0.509 Ku
0.399 Tu
0.132 Tu
0.496 Ku
0.374 Tu
0.123 Tu
0.480 Ku
0.315 Tu
0.112 Tu
0.430 Ku
0.242 Tu
0.084 Tu
Comment
= 0.585 , = 1 ,
N = 10, c = 1460
= 0.584 , = 1 ,
N = 10, c = 1400
= 0.577 , = 1 ,
N = 10, c = 1340
= 0.550 , = 1 ,
N = 10, c = 1250
= 0.543 , = 1 ,
N = 10, c = 1150
= 0.528 , = 1 ,
N = 10, c = 1050
= 0.504 , = 1 ,
N = 10, c = 930
= 0.495 , = 1 ,
N = 10, c = 840
= 0.484 , = 1 ,
N = 10, c = 730
= 0.477 , = 1 ,
N = 10, c = 630
= 0.454 , = 1 ,
N = 10, c = 520
= 0.445 , = 1 ,
N = 10, c = 410
= 0.433 , = 1 ,
N = 10, c = 300
= 0.385 , = 1 ,
N = 10, c = 190
= 0.286 , = 1 ,
N = 10, c = 100
= 0.158 , = 1 ,
N = 10, c = 6 0
Table 81: PID tuning rules - I 2PD model G m ( s) =
U(s) = Kc (1 +
K m e s m
- controller
s2
1
) E( s) + Kc ( b 1) R (s) Kc TdsY( s) . 1 tuning rule.
Tis
Rule
Kc
Direct synthesis
Hansen [91a]
Model: Method 1
3.75K m m
Ti
Td
Comment
5.4 m
2. 5 m
b = 0.167
Table 82: PID tuning rules SOSIPD model (repeated pole)
K m e s
s (1 + Tm1s) 2
1
Td s
U (s) = K c 1 +
+
Tis
T
1+ d
Rule
Kc
T
s
d
E( s) K c +
Td
s
1+
s
Ti
Servo/regulator
tuning
Taguchi and Araki
[61a]
Td
Comment
Kc
(105 d) 8
Ti
(105 d )
Td
m
1.0 ;
Tm
20% ; settling time
settling time of
et al. [10]
= 0. 601 , = 1 ,
(105d )
Model: Method 2

Pagola [134a]
1.672 K u
0.366 Tu
0.136 Tu
c = phase
K m = 1 ; Tm = 1 ;
0.1 < m < 10
1.236 K u
0.427 Tu
0.149 Tu
0.994 Ku
0.486 Tu
0.155 Tu
0.842 Ku
0.538 Tu
0.154 Tu
Model: Method 1
0.752 Ku
0.567 Tu
0.157 Tu
0.679 Ku
0.610 Tu
0.149 Tu
0.635 K u
0.637 Tu
0.142 Tu
0.590 Ku
0.669 Tu
0.133 Tu
0.551K u
0.690 Tu
0.114 Tu
N = 10, c = 1640
= 0.607 , = 1 ,
N = 10, c = 1600
= 0.610 , = 1 ,
N = 10, c = 1550
= 0.616 , = 1 ,
N = 10, c = 1500
= 0.605 , = 1 ,
N = 10, c = 1450
= 0.610 , = 1 ,
N = 10, c = 1400
= 0.612 , = 1 ,
N = 10, c = 1350
= 0.610 , = 1 ,
N = 10, c = 1300
= 0.616 , = 1 ,
Kc
Ti
(105d )
(105 d)
Td
(105 d)
1
0. 5667
=
0.1778 +
m
Kc
+ 0. 002325
Tm
N = 10, c = 1250
2
3
4
m
m
m
m
= Tm 0.2011 + 11.16
14. 98 + 13. 70 4.835
Tm
Tm
Tm
Tm

= Tm 1. 262 + 0.3620 m , = 0. 6666 , = 0. 8206 0.09750 m + 0.03845 m
Tm
Tm
Tm
Rule
Pagola [134a]
continued
Kc
Ti
Td
0.520 K u
0.776 Tu
0.087 Tu
0.509 Ku
0.810 Tu
0.068 Tu
Comment
= 0.609 , = 1 ,
N = 10, c = 1200
= 0. 611 , = 1 ,
N = 10, c = 1180
Table 83: PID tuning rules - SOSPD model with a negative zero
K m (1 sTm3 ) e s
(1 + sTm1 )(1 + sTm2 )
- controller with filtered
1
Td s
. 1 tuning rule.
derivative G c ( s) = Kc 1 +
+
Ti s 1 + Td s
N
Rule
Kc
Ti
Td
Tm1 + Tm2 Tm 3
K m( + m )
Tm1 + Tm2 Tm 3
Tm1 Tm2 (Tm1 + T m2 Tm3 )Tm 3

Tm1 + T m2 Tm3
N=10; = [ Tm1 , m ]
2Tm 2 Tm3
2Tm1 Tm 3
Tm2 1 (2 Tm1 Tm3 )Tm3
N=10; = [ Tm1 , m ]
Comment
Robust
Chien [50]
Model: Method 1
Km ( + m )
2 Tm1 Tm3
Tm1 > Tm 2 > Tm3
K m (1 sTm3 ) e s
(1 + sTm1 )(1 + sTm2 )
1 1 + Td s
. 1 tuning rule.
G c ( s) = Kc 1 +
Ti s 1 + Td s
N
Rule
Kc
Ti
Td
Comment
Tm 2
K m( + m )
Tm2
Tm1
Tm1 > Tm 2
Tm1
K m( + m )
Tm1
Tm2
Robust
Chien [50]
Model: Method 1
N=10; = [ Tm1 , m ]
Tm1 > Tm 2
N=10; = [ Tm1 , m ]
K m (1 sTm3 ) e s
(1 + sTm1 )(1 + sTm2 )
- series controller with
1
Td s
. 1 tuning rule.
1 +
Ts
Ts
i
1 + d
N
Rule
Kc
Ti
Td
Comment
Tm 2
K m( + m )
Tm2
Tm1 Tm3
Tm1 > Tm 2 > Tm3
Tm1
K m( + m )
Tm1
Tm 2 Tm3
Robust
Chien [50]
Model: Method 1
N=10; = [ Tm1 , m ]
Tm1 > Tm 2 > Tm3
N=10; = [ Tm1 , m ]
Table 86: PID tuning rules - SOSPD model with a positive zero
K m (1 + sTm 3 ) e s
(1 + sTm1 )(1 + sTm 2 )
- ideal controller
1
Gc ( s) = K c 1 +
T
s
i
Kc
Rule
Ti
( p0 q1 p1 )
Minimum IAE, ISE

2
p0
and ITAE Wang et
p 0 = Tm1 + Tm2 + Tm3 m
al. [97]
q 1 = Tm1 + Tm 2 + 05
. m
Model: Method 1
Td
Comment

p
p0 q 2 p 2 p1
q1 1 ,
,
p0
p 0 q1 p1 p 0
p2 = 05
. Tm1Tm 2 m
1
01
.
m
1
Tm1
p1 = Tm1Tm2 +
01
.
Tm 2
1
Tm 1
q 2 = Tm1Tm2
01
.
Tm 3
1
Tm1
0.5 m (Tm1 + Tm2 ) 0.5Tm3 m
+0.5 m ( Tm1 + Tm 2 )
= 35550
.
3.6167
Tm 2
Tm 1
m
T
T
0.4918 m 2 0.3318 m 3
0.4685
Tm 1
Tm1
Tm 1
= 39395
.
3.2164
Tm3
Tm1
m
T
T
.
0.3318 m 2 + 2.5356 m 3 , minimum IAE
14746
T
T
Tm1
m1
m1
m
Tm 2
Tm 3 m
m
Tm2
Tm3
+ 16185
.
58240
.
+
.
02383
.
+ 13508
.
10933
Tm1
Tm1
Tm1 Tm1
Tm1
Tm1
Tm1
Tm 2
m
T
T
0.6679 m 2 0.0564 m3
0.2383
Tm 1
Tm1
Tm1
Tm1
= 32950
.
34779
.
m
Tm 2
Tm 3

m
Tm2
Tm3
+ 21781
.
55203
.
+ m 14704
.
04685
.
+ 14746
.
Tm1
Tm1
Tm1 Tm1
Tm1
Tm1
Tm1
Tm3
m
T
T
.
0.0564 m2 + 2.5648 m3 , minimum ISE
13508
Tm1
Tm1
Tm1
Tm1
m
Tm2
Tm3

m
T
Tm3
+ 25336
.
55929
.
+ m 14407
.
0.5712 m2 + 15340
.
Tm1
Tm 1
Tm1 Tm 1
Tm1
Tm1
Tm1
Tm 2
m
T
T
0.3268 m2 0.5790 m 3
0.5712
Tm 1
Tm1
Tm1
Tm 1
Tm3
Tm1
m
Tm2
T
.
05790
.
+ 2.7129 m 3 , minimum ITAE
15340
Tm1
Tm1
Tm1
K m (1 + sTm 3 ) e s
(1 + sTm1 )(1 + sTm 2 )
- ideal controller with filtered
1
Td s
. 1 tuning rule.
derivative G c ( s) = Kc 1 +
+
Ti s 1 + Td s
N
Kc
Rule
Ti
Td
Comment
Robust
Tm1 + Tm2 +
Chien [50]
T m3 m
+ T m3 + m
K m ( + Tm 3 + m )
Tm3 m
Tm1 + Tm2 +
Tm3 m
+ Tm 3 + m
+ Tm 3 + m
+
Tm1
Model: Method 1
2Tm1 +
Tm3 m
+ Tm3 + m
K m ( + Tm 3 + m )
2 Tm1 +
Tm1 Tm2
Tm3 m
+ Tm2 +
+ Tm 3 + m
Tm3 m
Tm3 m
+ Tm3 + m
+ Tm 3 + m
+
Tm2 3
Tm 3 m
2 Tm1 +
+ Tm3 + m
Tm1 > Tm 2 > Tm3
N=10; = [ Tm1 , m ]
Tm1 > Tm 2 > Tm3
N=10; = [ Tm1 , m ]
K m (1 + sTm 3 ) e s
(1 + sTm1 )(1 + sTm 2 )
1 1 + Td s
. 1 tuning rule.
G c ( s) = Kc 1 +
Ti s 1 + Td s
N
Rule
Kc
Ti
Td
Comment
Tm 2
K m ( + Tm3 + m )
Tm2
Tm1
Tm1 > Tm 2 > Tm3
Tm1
K m ( + Tm3 + m )
Tm1
Tm2
Robust
Chien [50]
Model: Method 1
N=10; = [ Tm1 , m ]
Tm1 > Tm 2 > Tm3
N=10; = [ Tm1 , m ]
K m (1 + sTm 3 ) e s
(1 + sTm1 )(1 + sTm 2 )
- series controller with
1
Td s
. 1 tuning rule.
1 +
Ts
Ts
i
1 + d
N
Rule
Kc
Ti
Tm 2
K m ( + Tm3 + m )
Tm2
Tm1
K m ( + Tm3 + m )
Tm1
Td
Robust
Chien [50]
Model: Method 1
Tm1 +
Tm 3 m
+ Tm 3 + m
Tm 2 +
Tm 3 m
+ Tm3 + m
Comment
N=10; = [ T, m ] , T =
dominant time
constant
N=10; = [ T, m ] , T =
dominant time
constant
Km e s
- ideal controller
(1 + sTm1 )(1 + sTm 2 )(1 + sTm 3 )
m
Table 90: PID tuning rules - TOLPD model
1
Gc ( s) = K c 1 +
Ti s
Rule
Kc
Ti
Td
Comment

Standard form
optimisation binomial - Polonyi
[153]
Model: Method 1
Standard form
optimisation
minimum ITAE Polonyi [153]
Model: Method 1
1 4 Tm 2 + Tm2 Tm1
T
m
m3 m
4 6Tm2 m + m
Tm1 > 10( Tm2 + m )
Tm2
T T
1 21
.
+ m2 m1
3
.
4
Tm3 m
m
2.7 3.4Tm2 m + m
Tm1 > 10( Tm2 + m )
Table 91: PID tuning rules - TOLPD model - G m (s ) =
K m e s
(1 + sTm ) 3
1
Td s
U (s) = K c 1 +
+
Tis
T
1+ d
Kc
Rule
Ti
Servo/regulator
tuning
Kc
(105 e ) 2
Ti
Kc
(105e )
Td
Comment
Taguchi and Araki

[61a]
T
s
d
E( s) K c +
R (s) . 1 tuning rule.
Td
s
1+
s
1
1.275
=
0.4020 +
m
Kc
0.003273
Tm
(105 e )
Td
(105 e )
m
1.0 ;
Tm
20% ; settling time
settling time of
et al. [10]
2
3
4
m
m
m
m
Ti
= Tm 0. 3572 + 7. 647
12. 86 + 11 .77 4. 146
Tm
Tm
Tm
Tm
2
2
m
m
m
m
(105 e )
Td
= Tm 0.8335 + 0.2910
0. 04000 , = 0. 6661 0.2509
+ 0. 04773 ,
Tm
Tm
Tm
Tm
(105 e )
= 0. 8131 0.2303
m

+ 0.03621 m
Tm
Tm
K m e s
1
- ideal controller Gc ( s) = K c 1 +
+ Td s .
1 sTm
Ti s
Table 92: PID tuning rules - unstable FOLPD model
3 tuning rules.
Kc
Rule
Ti
Td
Comment
Direct synthesis
De Paor and
OMalley [86]
Tm
cos
K m
(1 T )T
Tm 1 Tm m
sin
Km
Tm m
m m
(1 T )T
m
Model: Method 1
Chidambaram [88]
Model: Method 1
Valentine and
Chidambaram [154] dominant pole
placement
Model: Method 1
1
T
. + 0.3 m
13
Km
m
Tm
1165
.

Km m
0.245
1 T
m m
Tm
Tm m
m = tan 1
0.245
Tm
1165
.

Km m
0.245
1 Tm m
2
Tm 1 Tm m
1 Tm m
Tm m 1 Tm m
Tm m

Tm 25 27 m
Tm
m
0176
T
.
+ 0.36
Tm m
0.179 0.324
Tm
1165
.

Km m
tan 0.75
m
m
Tm
+ 0.161
Tm
Ti
046
. m
m
< 0.6
Tm

.
+ 0.36 m Tm
0176
Tm
m
< 06
.
Tm

.
+ 0.36 m 25Tm 0176
.
+ 0.36 m Tm
0176
Tm
Tm
0.176Tm + 0.36 m
0.12 0.1 m
Tm
m
<1
Tm

.
+ 0.36 m Tm
0176
Tm
0.6
08
. <
m
Tm
m
Tm
08
.
K m e s
1 sTm
m
1
KTs
U( s) = Kc 1 + E ( s) c d Y(s) . 2 tuning rules.
sT
Ti s
1+ d
N
Kc
Rule
Servo tuning
Huang and Lin [155]
- minimum IAE
Model: Method 2
Regulator tuning
Huang and Lin [155]
- minimum IAE
Model: Method 2
Ti
Td
Comment
1 .0251
m
m
1
0 .433 + 0 .2056
+ 0 .3135
Km
Tm
T m
6.6423
7.6983 m

m
T 0. 0312+ 1. 6333 m + 00399
Tm 00018
.
+ 08193
.
+ 7. 7853 m
.
e T m
m
Tm
Tm
Tm
0.01
m
0.8 ; N=10
Tm
0.01
m
0.8 ; N=10
Tm
1.004
1
m

0.2675 + 01226
.
+ 0.8781 m
Km
Tm
Tm
2. 9123
Tm 00005
.
+ 2.4631 m + 95795
.

T
T
m

Tm 0.0011 + 0.4759 m
Tm
K m e s
1 sTm
m
1 1 + Td s
. 1 tuning rule.
G c ( s) = Kc 1 +
Ti s 1 + Td s
N
Rule
Kc
Regulator tuning
Shinskey [16] minimum IAE page
381.
Method: Model 1
Ti
Td
Comment

170
. m
0.60m
m Tm = 01
.
Tm
K m m
1.90 m
0.60m
m Tm = 0.2
08929
.
Tm
Km m
2.00 m
0.80 m
m Tm = 0.5
08621
.
Tm
K m m
2.25m
0.90m
m Tm = 0.67
08333
.
Tm
K m m
2.40 m
100
. m
m Tm = 0.8
0.9091Tm
Km m
Table 95: PID tuning rules - unstable SOSPD model G m ( s) =
Km e s
- ideal controller
(1 sTm1 )(1 + sTm 2 )
m
1
Gc ( s) = K c 1 +
Ti s
Kc
Ti
Td
Comment
Kc(106 ) 3
Ti (106)
Td (106 )
Tuning rules
Tm1
+ 2 + Tm 2
Tm1
+ 2 + Tm 2
Tm1
m Tm = 02
.
= [0.5Tm ,19
. Tm ]
m Tm = 0.4
= [1.3Tm ,19
. Tm ]
m Tm = 02
.
= [0.4Tm ,4. 3Tm ]
m Tm = 0.4
= [11
. Tm ,4.3Tm ]
m Tm = 0.6
= [ 2.2Tm , 4.3Tm ]
Rule
Ultimate cycle
McMillan [58]
Model: Method not
relevant
Robust
Rotstein and Lewin
[89]
Model: Method 1
+ 2 Tm 2
Tm1
+ 2 + Tm 2
Tm1
determined
graphically sample
values provided
Kc
Ti
( 106)
( 106)
1111
.
Tm1Tm2
Km m 2
0.65 ,

( Tm1 + Tm 2 )Tm1Tm2
1 + T T T
m2 )( m1
m) m

( m1
0.65
0.65
(T + T ) T T
(T + T )T T

m1
m2
m1 m2
( 106)
m1
m2
m1 m 2
= 2 m 1 +
= 05
. m 1 +
, Td

( Tm1 Tm2 )( Tm1 m ) m
(Tm1 Tm 2 )( Tm1 m ) m
Km e s
(1 sTm1 )(1 + sTm 2 )
m
1 1 + Tds
. 2 tuning rules.
G c ( s) = Kc 1 +
Ti s 1 + Td s
N
Kc
Rule
Ti
Regulator tuning
bTm1 1 +
aTm2 2
Minimum ITAE 2
4( m + Tm 2 )
Poulin and
Pomerleau [82], [92] K m ( aTm1 4[ m + Tm2 ])
deduced from graph
Model: Method 1
( m + Td N) Tm1
Output step load

disturbance
Input step load

disturbance
Td
Comment
4Tm1 ( m + Tm2 )
aTm1 4( m + Tm2 )
Tm2
m
2;
T
( d N)
01
. Tm
Td
033
. Tm
N
a
0.9479
1.0799
1.2013
1.3485
1.4905
b
2.3546
2.4111
2.4646
2.5318
2.5992
( m + Td N) Tm1
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
a
1.6163
1.7650
1.9139
2.0658
2.2080
b
2.6612
2.7368
2.8161
2.9004
2.9826
0.05
0.10
0.15
0.20
0.25
1.1075
1.2013
1.3132
1.4384
1.5698
2.4230
2.4646
2.5154
2.5742
2.6381
0.30
0.35
0.40
0.45
0.50
1.6943
1.8161
1.9658
2.1022
2.2379
2.7007
2.7637
2.8445
2.9210
3.0003
Km e s
- series controller
(1 sTm1 )(1 + sTm 2 )
m
1
Gc ( s) = Kc 1 +
(1 + Td s) . 1 tuning rule.
Ti s
Rule
Kc
Direct synthesis
Ho and Xu [90]
p Tm1
Model: Method 1
A m Km
Ti
Td
157
. p p m
2
1
Tm1
Tm2
Comment
p =
A m m + 157
. A m (A m 1)
(A
2
m
1 m
Km e s
(1 sTm1 )(1 + sTm 2 )
m
1
KTs
U( s) = Kc 1 + E ( s) c d Y(s) . 2 tuning rules.
sT
Ti s
1+ d
N
Rule
Servo tuning
Huang and Lin [155]
- minimum IAE
Kc
Ti
Td
Comment
Tm2 Tm1 ;
Kc(107 ) 4
Ti (107)
Td (107 )
0.05
Model: Method 2
Kc
( 107 )
m
0.4
Tm1
1.344
0.995
m
Tm 2
1
m
Tm2
T
=
10.741 13363
.
+ 0099
.
+ 727.914
708.481
+ 9.915 m 2 2m
Km
Tm1
Tm1
Tm1
Tm1
Tm1
1.031
0.997
T
T
T
Tm 2
Tm 2
1
T
T
T
84.273 m1
90
.
959
+
9
.
034
e
2
.
386
e
16
.
304
e
Km
m
Tm1
m
m2
m1
m1
m 2 m
2
m1
2.12
0.985

T
m
T
T
Ti (107 ) = Tm1 149 .685 141418
.
88.717 m 17.29 m 2 + 20518
. m2
12.82 m 2 2m
Tm1
Tm1
Tm1
Tm1
Tm1
0.286
1.988
T
T

Tm 2
T
+ Tm1 3.611 m
+ 0.000805 m2 + 141.702e T 2.032e T + 10.006e T
Tm1
Tm1
m
Td
( 107 )
m2
m1
m1
m 2 m
2
m1
2
2
m
Tm2
m
Tm2
T
= Tm1 0.4144 + 15805
.
142.327
+ 01123
.
+ 0.7287
18.317 m 2 2m
Tm1
Tm1
Tm1
Tm1
Tm1
3
3
2T
T 2

T
+ Tm1 48695
. m 10542
. m2 + 204.009 m 3m 2 + 47.26 m m32
Tm1
Tm1
Tm1
Tm1

+ 396349
. m
Tm1
3T
T 3
2T 2
T

+ Tm1 138.038 m 4m 2 + 52155
. m m42 646.848 m m4 2 + 19.302 m 2 4731.72 m
Tm1
Tm1
Tm1
Tm1
Tm1
4
4T
T 4
3T 2
+ Tm1 425789
. m 5m 2 289 .746 m m52 841807
. m m5 2
Tm1
Tm1
Tm1
2T 3
Tm 2
+ 1313.72 m m5 2 37688
.
Tm1
Tm1
5T
T 5
4T 2
m
+ Tm1 6264 .79
. m 6m2 + 204.689 m m62 + 25706
. m m6 2
161469
Tm1
Tm1
Tm1
Tm1
3
6
2T 4
T
T
+ Tm1 791857
. m m6 2 + 648.217 m m2 2 5.71 m2
Tm1
Tm1
Tm1
Rule
Kc
Ti
Td
Huang and Lin [155]

- minimum IAE continued
Model: Method 2
Kc(108 ) 5
Ti (108)
Td (108 )
Kc (108) =
Ti
Tm 1 < Tm 2 10Tm1 ;
0.05
m
025
.
Tm1
0.3055
0.5174

T
1
m
T
T
1302
. + 85914
.
+ 34.82 m
+ 10.442 m 2 22.547 m2
14.698 m 2 2m
Km
Tm1
Tm1
Tm1
Tm1
Tm2
Km
( 108)
Comment
1.0077
0.9879
T
Tm 2
Tm 2
T
52.408 m1
5147
.
+ 53.378e 0.000001e T
m
m1
m
m2
m1
m1
Tm2 m
+ 0.286e
Tm1 2
2
2
Tm 2
m
Tm2
m
Tm2 m
= Tm1 72.806 268.746
4.9221
.
.
+ 246819
+ 0.6724
+ 151351
Tm1
Tm1
Tm1
Tm1
Tm12
3
3
4
2T
T 2

Tm 2

+ Tm1 6914.46 m 00092
.
. m 3m 2 14.27 m m32 + 558017
. m
795465
Tm1
Tm1
Tm1
Tm1
Tm1
3T
T 3
2T 2
Tm2
+ Tm 1 1417 .65 m 4m 2 + 0.4057 m m42 + 55536
. m m4 2 0001119
.
Tm1
Tm1
Tm1
Tm1
T
T
T
+ Tm 1 44.903e + 0.000034e 15694
. eT
m2
m1
m1
m2 m
2
m1

+ 678778 m
Tm1
19.056
Tm2
Tm1
7.3464
1.1798
0.1064
T
T
Td (108) = Tm1 175515
.
86.2 m + 348.727 m
0.008207 m 2 55619
. m2
+ 0.0418 m 2 2m
Tm1
Tm1
Tm1
Tm1
Tm1
0.0355
0.0827

Tm2
T
+ Tm 1 78959
. m
+ 0.005048 m 2 187 .01e T
Tm 1
Tm1
m
Tm 2 m
Tm 2
m1
+ 0000001
.
e
Tm1
00149
.
e
Tm1 2
Kc
Rule
Regulator tuning
Huang and Lin [155]
minimum IAE
Ti
Td
Comment
Tm2 Tm1 ;
Kc
( 109 ) 6
Ti
( 109)
Td
( 109 )
0.05
Model: Method 2
Kc
( 109 )
1.164
2.54
m
Tm2
1
m
Tm 2
T
=
174167
.
31364
.
+ 0.4642
103069
.
83916
.
66.962 m2 2m
Km
Tm1
Tm1
Tm1
Tm1
Tm1
Km
Ti
( 109)
m
0.4
Tm1
1.065
1.014
T T
Tm 2
T
59.496 m1 m 2
7079
.
+ 23121
. e
m
m Tm1
Tm 2 m
Tm 2
m1
+ 126.924e
Tm 1
+ 26.944e
Tm 1 2
2
2
3
m
Tm2
m
m
Tm 2
Tm2 m
= Tm1 0.008 + 2.0718
+ 6.431
+ 0.7503
18686
.
+ 0.4556
+ 2.4484

Tm1
Tm1
Tm1
Tm1
Tm12
Tm1
3
4
T 2
T 2
T 3
T

+ Tm 1 2.9978 m 2 21135
. m2 m3 + 12.822 m m32 + 39.001 m + 22848
. m 2 4m
Tm1
Tm1
Tm1
Tm1
Tm1
T 3
T 2 2
T
+ Tm1 4.754 m 2 4 m 0.527 m2 4m + 164
. m2
Tm1
Tm1
Tm1
2
2
3

Tm2

m
Tm 2
T
Td (109) = Tm1 0.0301 + 11766
.
4.4623 m + 05284
.
14281
.
. m
+ 4.6 m 2 2m + 11176
Tm1
Tm1
Tm1
Tm1
Tm1
Tm1
3
4
m2 Tm 2
m Tm2 2
Tm 2 m 3
Tm 2
m
+ Tm 1 10886
.
05039
.
98564
.
7528
.
5.0229
3
3
4
Tm1
Tm1
Tm1
Tm1
Tm1
4
T 3
2T 2
Tm 2
+ Tm1 2.3542 m m42 + 9.3804 m m4 2 01457
.
Tm1
Tm1
Tm1
Rule
Kc
Ti
Td
Huang and Lin [155]

- minimum IAE
continued
Model: Method 2
Kc(110 ) 7
Ti (110)
Td (110 )
Kc
( 110)
Comment
Tm 1 < Tm 2 10Tm1 ;
0.05
m
025
.
Tm1
2.1984
0.791
m
Tm 2
1
m
Tm 2
T
=
1750.08 + 1637.76
+ 1533.91
7.917
+ 6187
.
6.451 m2 2m
Km
Tm1
Tm1
Tm1
Tm1
Tm1
3.2927
1.0757
T
Tm1
Tm 2
1
Tm 2
T
0002452
+ 13729
.
1739.77e 0.000296e T
Km
m
m
Tm1
m2
m1
m1
Tm 2 m
+ 2.311e
Tm1 2
2
2
3

Tm 2

Tm 2
T
Ti (110) = Tm1 51678
.
57.043 m + 1337.29 m + 01742
.
01524
.
+ 7.7266 m 2 2m 6011.57 m
Tm1
Tm1
Tm1
Tm1
Tm1
Tm1
3
4
Tm2 m 2
m Tm2 2
Tm2 m 3
Tm2
m
+ Tm 1 0.0213
+
0
.
0645
+
913552
.
+
274
.
851
65.283
3
3
4
Tm1
Tm1
Tm1
Tm1
Tm1
4
T 3
T 2 2
Tm 2
T
T
+ Tm 1 0.003926 m 2 4 m 2.0997 m 2 4m 0001077
.
49
.
007
e
+
0
.
000026
e
T
T
T
m1
m1
m1
T
+ Tm1 0.2977e T
m2
m1
m2
m1
m1
2
2
3
T
T
Td (110) = Tm1 0.0605 + 4.6998 m 29.478 m + 0.0117 m 2 0.0129 m 2 + 0.6874 m2 2m + 140.135 m
Tm1
Tm1
Tm1
Tm1
Tm1
Tm1
3
4
m 2 Tm 2
mTm 2 2
Tm 2 m 3
Tm 2
m
+ Tm 1 0.002455
.
01289
.
238
.
864
+
06884
.
14712
3
3
4
Tm1
Tm1
Tm1
Tm1
Tm1
4
T 3
m 2Tm 2 2
Tm 2
+ Tm 1 0.007725 m m42 01222
.
0
.
000135
Tm1
Tm1
Tm1
Table 99: PID tuning rules general model with a repeated pole G m ( s) =
K m e s
- ideal controller
(1 + sTm ) n
m
1
Gc ( s) = K c 1 +
Ti s
Kc
Rule
Direct synthesis
Skoczowski and
Tarasiejski [156]
Model: Method 1
g Tm
Km
(1 +
2
g
Tm 2
Ti
Td
Tm
n1
Tm , n 2
n+2
Comment
n 1
1 + g 2 Td 2
1
2n + 4 m 4n + 2
Tm n 2 2n 2 +
+
m b
Tm
g =
2
4 n + 2 m 2n 2
2
2Tm n 4n + 3 +
+
m
Tm
with
2
b = Tm
2
2n + 4 m 4 n + 2 2
2
4 n + 2 m 2 n 2
+
m + 4( n + 2) 1 m n2 4n + 3 +
+
m
n 2n 2 +
Tm

Tm
Table 100: PID tuning rules general stable non-oscillating model with a time delay - ideal controller
1
Gc ( s) = K c 1 +
Ti s
Kc
Rule
Ti
Td
Comment
Direct synthesis
Kc
(110 a )
Ti
Ti
(110 a ) 2
Ti
(110 a )
Kc
(110 a )
Ti
(110 a )
Kc
(110 a )
Ti
(110 a )
Kc
Gorez and Klan

[147a]
Model: not specified
( 110a )
( 110a )
+ m
, Ti
(110 a )
1+ 1 + 2 m
Tar
= Tar
2
Td
Ti
(110 a )
(110a )
m
Tar
0.25Ti
m
1
T
ar
(110 a )
2 m
Tar
2 m
Ti (110a ) + Tcr cr + Ti (110a ) Taa m 1 K c (110a ) 1 +

(110 a )
Tar
2Tar
3Ti (110 a )
Ti
Tar = average residence time of the process (which equals Tm + m for a FOLPD process, for example); T aa =
Td
(110 a )
Tar
(110 a )
1 + m
additional apparent time constant; Tcr = 1 K c
2T (110a )
Table 101: PID tuning rules fifth order model with delay
1
G m ( s) =
ideal controller Gc ( s) = K c 1 +
+ Td s .
2
3
4
5
Ti s
1 + a 1s + a 2 s + a 3s + a 4 s + a 5s
1 tuning rule.
Kc
Rule
Direct synthesis
Magnitude optimum
- Vrancic et al. [159]
Model: Method 1
Ti
Td
Ti (112)
Td (112 )
Comment
Kc
( 112 )
Kc ( 112 ) =
Td
( 112)
2 Km a1 b 1 + a1a2 + a1b1 a3 b1b 2 + b 3 + ( a1 b 1 ) m + ( a1 b 1 ) m2 + 0.333 m 3 ( a1 b 1 + m ) Td

2
a1 a1 b 1 + a 1b 2 2 a1a 2 + a2 b 1 + a 3 b 3 + m a 1 a1b 1 a 2 + b 2 + 0.5( a1 b1 ) m + 0.167 m

3
Ti (112 ) =
a13 a1 2b 1 + a1 b 2 2 a1a2 + a2 b 1 + a 3 b 3 + m a1 2 a1b1 a 2 + b 2 + 0.5( a1 b1 ) m 2 + 0167

. m 3
[a
2
1
a 1b 1 a 2 + b 2 + ( a1 b 1 ) m + 0.5 m ( a1 b1 + m ) Td
see attached sheet with = 0.
Table 102: PID tuning rules fifth order model with delay
G m ( s) =
controller with filtered derivative
1 + a1s + a2 s2 + a3s 3 + a 4 s4 + a 5s5
m
1
Tds
. 1 tuning rule.
G c ( s) = Kc 1 +
+
Ts
Ts
i
1 + d
N
Rule
Direct synthesis
Magnitude optimum
- Vrancic et al. [73]
Model: Method 1
Kc
Ti
Td
Comment
Kc(113) 4
Ti (113)
Td (113)
8 N 20
Kc ( 113) =
Ti( 113 ) =
a13 a12 b1 + a1b2 2a1a2 + a 2b 1 + a 3 b 3 + m a12 a1b 1 a2 + b 2 + 0.5( a1 b1 ) m2 + 0.167 m3

2
T2
2
2
2
2
3
2 K m a1 b 1 + a1a2 + a1b 1 a3 b1b 2 + b 3 + ( a1 b1) m + ( a1 b1 ) m + 0.333 m ( a1 b1 + m ) Td d ( a1 b1 + m )
N
a13 a12 b1 + a1b 2 2a1a2 + a2 b1 + a 3 b 3 + m a12 a1b 1 a 2 + b 2 + 0.5( a1 b1 ) m2 + 0.167 m3

2
Td 2
2
a1 a1b 1 a 2 + b 2 + ( a1 b 1) m + 0.5 m ( a1 b1 + m ) Td
Td (113) = see attached sheet
4. Conclusions
The report has presented a comprehensive summary of the tuning rules for PI and PID controllers that
have been developed to compensate SISO processes with time delay. Further work will concentrate on evaluating
the applicability of these tuning rules to the compensation of processes with time delays, as the value of the time
delay varies compared with the other dynamic variables.
Year
1942
1950
1951
1952
1953
1954
1961
1964
1965
1967
1968
1969
1972
1973
1975
1979
1980
1982
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
1941-1950
Table 103: Tuning rules published by year and medium

Journal articles
Conference
Books/ Ph.D.
papers/
thesis
correspondence
1
1
1
1
3
1
1
1
1
1
1
1
2
2
1
1
2
1
2
1
2
1
1
1
1
1
3
4
2
4
3
4
1
5
1
2
5
4
1
5
2
2
14
2
1
8
6
2
11
4
1
6
7
12
1
1
1
16
2
Total
1
1
1
1
3
1
1
1
1
2
1
2
2
2
2
1
2
1
3
2
1
1
9
7
5
5
3
10
9
17
16
16
13
14
17
2
1951-1960
1961-1970
1971-1980
1981-1990
1991-2000
TOTAL
6
5
7
15
68
103
0
0
0
7
44
51
0
3
2
7
8
20
6
8
9
29
120
174
List of journals in which tuning rules were published and number of tuning rules published
Advances in Modelling and Analysis C, ASME Press
1
AIChE Journal
4
Automatica
12
British Chemical Engineering
1
Chemical Engineering Communications
5
Chemical Engineering Progress
1
Chemical Engineering Science
1
Control
1
Control and Computers
1
Control Engineering
7
Control Engineering Practice
4
European Journal of Control
1
Hydrocarbon Processing
2
Hungarian Journal of Industrial Chemistry
1
IEE Proceedings, Part D
9
(including IEE Proceedings - Control Theory and Applications, Proceedings of the IEE, Part 2)
IEEE Control Systems Magazine
1
IEEE Transactions on Control Systems Technology
4
IEEE Transactions on Industrial Electronics and Control Instrumentation
1
IEEE Transactions on Industrial Electronics
1
IEEE Transactions on Industry Applications
1
Industrial and Engineering Chemistry Process Design and Development 2
Industrial Engineering Chemistry Research
12
International Journal of Adaptive Control and Signal Processing
1
International Journal of Control
3
International Journal of Electrical Engineering Education
1
International Journal of Systems Science
1
Instrumentation
1
Instrumentation Technology
2
Instruments and Control Systems
2
ISA Transactions
2
Journal of Chemical Engineering of Japan
2
Journal of the Chinese Institute of Chemical Engineers
3
Pulp and Paper Canada
1
Process Control and Quality
1
Thermal Engineering (Russia)
2
Transactions of the ASME
7
(including Transactions of the ASME. Journal of Dynamic Systems, Measurement and Control)
Transactions of the Institute of Chemical Engineers
1
Classification of journals in which tuning rules were published and number of tuning rules
published
Chemical Engineering Journals
32
(AIChE Journal, British Chemical Engineering, Chemical Engineering Communications, Chemical Engineering
Progress, Chemical Engineering Science, Transactions of the Institute of Chemical Engineers, Hungarian Journal
of Industrial Chemistry, Industrial and Engineering Chemistry Process Design and Development, Industrial
Engineering Chemistry Research, Journal of Chemical Engineering of Japan, Journal of the Chinese Institute of
Chemical Engineers)
Control Engineering Journals
53
(Automatica, Control, Control and Computers, Control Engineering, Control Engineering Practice, European
Journal of Control, IEE Proceedings, Part D, IEE Proceedings - Control Theory and Applications, Proceedings of
the IEE, Part 2, IEEE Control Systems Magazine, IEEE Transactions on Control Systems Technology,
International Journal of Adaptive Control and Signal Processing, International Journal of Control, International
Journal of Systems Science, Instrumentation, Instrumentation Technology, Instruments and Control Systems,
ISA Transactions, Process Control and Quality)
Mechanical Engineering Journals
8
(Advances in Modelling and Analysis C, ASME Press, Transactions of the ASME, Transactions of the ASME.
Journal of Dynamic Systems, Measurement and Control)
Electrical/Electronic Engineering Journals
4
(EE Transactions on Industrial Electronics and Control Instrumentation, IEEE Transactions on Industrial
Electronics, IEEE Transactions on Industry Applications, International Journal of Electrical Engineering
Education)
Trade Journals
5
(Hydrocarbon Processing, Pulp and Paper Canada, Thermal Engineering (Russia))
5. References
1.
Koivo, H.N. and Tanttu, J.T., Tuning of PID Controllers: Survey of SISO and MIMO techniques.
Proceedings of the IFAC Intelligent Tuning and Adaptive Control Symposium, Singapore, 1991, 75-80.
2.
Hwang, S.-H., Adaptive dominant pole design of PID controllers based on a single closed-loop test.
Chemical Engineering Communications, 1993, 124, 131-152.
3.
Astrom, K.J. and Hagglund, T., PID Controllers: Theory, Design and Tuning. Instrument Society of America,
Research Triangle Park, North Carolina, 2nd Edition, 1995.
4.
Bialkowski, W.L., Control of the pulp and paper making process. The Control Handbook. Editor: W.S.
Levine, CRC/IEEE Press, Boca Raton, Florida, 1996, 1219-1242.
5.
Isermann, R., Digital Control Systems Volume 1. Fundamentals, Deterministic Control. 2nd Revised Edition,
Springer-Verlag, 1989.
6.
Ender, D.B., Process control performance: not as good as you think. Control Engineering, 1993, September,
180-190.
7.
Astrom, K.J. and Wittenmark, B., Computer controlled systems: theory and design. Prentice-Hall
International Inc., 1984.
8.
Ziegler, J.G. and Nichols, N.B., Optimum settings for automatic controllers. Transactions of the ASME, 1942,
November, 759-768.
9.
Hazebroek, P. and Van der Waerden, B.L., The optimum adjustment of regulators. Transactions of the
ASME, 1950, April, 317-322.
10. Chien, K.-L., Hrones, J.A. and Reswick, J.B., On the automatic control of generalised passive systems.
Transactions of the ASME, 1952, February, 175-185.
11. Cohen, G.H. and Coon, G.A., Theoritical considerations of retarded control. Transactions of the ASME, 1953,
May, 827-834.
12. Wolfe, W.A., Controller settings for optimum control. Transactions of the ASME, 1951, May, 413-418.
13. Murrill, P.W., Automatic control of processes. International Textbook Co., 1967.
14. McMillan, G.K., Tuning and control loop performance - a practitioners guide. Instrument Society of
America, Research Triangle Park, North Carolina, 3rd Edition, 1994.
15. St. Clair, D.W., Controller tuning and control loop performance, Straight Line Control Co., Inc., 2nd Edition,
1997.
15a. Shinskey, F.G. (2000). PID-deadtime control of distributed processes, Preprints of the Proceedings of PID
00: IFAC Workshop on digital control (Past, present and future of PID control), Terrassa, Spain, pp. 1418.
16. Shinskey, F.G., Process Control Systems - Application, Design and Tuning. McGraw-Hill Inc., New York, 3rd
Edition, 1988.
17. Shinskey, F.G., Process Control Systems - Application, Design and Tuning. McGraw-Hill Inc., New York, 4th
Edition, 1996.
18. Huang, C.-T., Chou, C.-J. and Wang, J.-L., Tuning of PID controllers based on second-order model by
calculation. Journal of the Chinese Institute of Chemical Engineers, 1996, 27(2), 106-120.
19. Yu, S. W., Optimal PI tuning for load disturbances. Journal of the Chinese Institute of Chemical Engineers,
1988, 19(6), 349-357.
20. Zhuang, M. and Atherton, D.P., Automatic tuning of optimum PID controllers. IEE Proceedings, Part D, 1993,
140(3), 216-224.
21. Rovira, A.A., Murrill, P.W. and Smith, C.L., Tuning controllers for setpoint changes. Instruments and Control
Systems, 1969, 42, December, 67-69.
22. Khan, B.Z. and Lehman, B., Setpoint PI controllers for systems with large normalised dead time. IEEE
Transactions on Control Systems Technology, 1996, 4(4), 459-466.
23. Haalman, A., Adjusting controllers for a deadtime process. Control Engineering, 1965, July, 71-73.
23a. Chen, C.-L. and Yang, S.-F., PI tuning based on peak amplitude ratio, Preprints of the Proceedings of PID
00: IFAC Workshop on digital control (Past, present and future of PID control), Terrassa, Spain, 2000, pp.
195-198.
24. Pemberton, T.J., PID: The logical control algorithm. Control Engineering, 1972, 19(5), 66-67.
25. Smith, C.A. and Corripio, A.B., Principles and practice of automatic process control. John Wiley and Sons,
New York, 2nd Edition, 1997.
26. Smith, C.L., Corripio, A.B. and Martin, J. (1975). Controller tuning from simple process models,
Instrumentation Technology, December, 39-44.
27. Hang, C.C., Tan, C.H. and Chan, W.P., A performance study of control systems with dead time. IEEE
Transactions on Industrial Electronics and Control Instrumentation, 1980, IECI-27(3), 234-241.
27a. Miluse, V., Vitecek, A. and Smutny, L. (2000a). Controller tuning for controlled plants with time delay,
Preprints of the Proceedings of PID 00: IFAC Workshop on digital control (Past, present and future of
PID control), Terrassa, Spain, pp. 283-288.
27b. Miluse, V., Vitecek, A. and Smutny, L. (2000a). Controller tuning for controlled plants with time delay,
PID control), Terrassa, Spain, pp. 289-294.
28. Gorecki, H., Fuska, S., Grabowski, P. and Korytowski, A., Analysis and synthesis of time delay systems,
John Wiley and Sons, New York, 1989.
29. Chiu, K.C., Corripio, A.B. and Smith, C.L., Digital controller algorithms. Part III. Tuning PI and PID
controllers. Instruments and Control Systems, 1973, December, 41-43.
30. Astrom, K.J., Hagglund, T., Hang, C.C. and Ho., W.K., Automatic tuning and adaptation for PID controllers a survey. Control Engineering Practice, 1993, 1(4), 699-714.
31. Davydov, N.I., Idzon, O.M. and Simonova, O.V., Determining the parameters of PID-controller settings using
the transient response of the controlled plant. Thermal Engineering, 1995, 42(10), 801-807.
32. Schneider, D.M., Control of processes with time delay. IEEE Transactions on Industry Applications, 1988, 24
(2), 186-191.
33. McAnany, D.E., A pole placement technique for optimum PID control parameters. Proceedings of the ISA/93
Advances in Instrumentation and Control Conference, McCormick Place, Chicago, Illinois, USA, 1993, 48,
1775-1782.
34. Leva, A., Maffezzoni, C. and Scattolini, R., Self-tuning PI-PID regulators for stable systems with varying
delay. Automatica, 1994, 30(7), 1171-1183.
35. Hang, C.C., Ho, W.K. and Cao, L.S., A comparison of two design methods for PID controllers. Proceedings
of the ISA/93 Advances in Instrumentation and Control Conference, McCormick Place, Chicago, Illinois,
USA, 1993, 48, 959-967.
36. Hang, C.C., Lee, T.H. and Ho, W.K., Adaptive Control. Instrument Society of America, Reseaech Triangle
Park, North Carolina, 1993.
37. Ho, W.K., Hang, C.C. and Zhou, J.H., Tuning of PID controllers based on gain and phase margin
specifications. Automatica, 1995, 31(3), 497-502.
38. Kookos, I.K., Lygeros, A.I. and Arvanitis, K.G., On-line PI controller tuning for integrator/dead time
processes. European Journal of Control, 1999, 5, 19-31.
39. Tan, K.K., Lee, T.H. and Wang, Q.G., Enhanced automatic tuning procedure for process control of PI/PID
controllers. AIChE Journal, 1996, 42(9), 2555-2562.
40. Voda, A. and Landau, I.D., The autocalibration of PI controllers based on two frequency measurements.
International Journal of Adaptive Control and Signal Processing, 1995, 9, 395-421.
41. Friman, M. and Waller, K.V., A two channel relay for autotuning. Industrial Engineering Chemistry Research,
1997, 36(7), 2662-2671.
42. Smith, L., A modified Smith predictor for extruded diameter control. InstMC Mini Symposium - Algorithms
and Architectures for Industrial Controllers (in UKACC International Conference on Control 98), Swansea,
Wales, Lecture 5, 1998.
43. Cox, C.S., Daniel, P.R. and Lowdon, A., Quicktune: a reliable automatic strategy for determining PI and PPI
controller parameters using a FOLPD model. Control Engineering Practice, 1997, 5(10), 1463-1472.
44. Cluett, W.R. and Wang, L., New tuning rules for PID control. Pulp and Paper Canada, 1997, 3(6), 52-55.
45. Abbas, A., A new set of controller tuning relations. ISA Transactions, 1997, 36(3), 183-187.
46. Bi, Q., Cai, W.-J., Lee, E.-L., Wang, Q.-G., Hang, C.-C. and Zhang, Y., Robust identification of first-order plus
dead-time model from step response, Control Engineering Practice, 1999, 7(1), 71-77.
47. Wang, Y.-G. and Shao, H.-H. (2000). Optimal tuning for PI controller, Automatica, 36, 147-152.
48. Brambilla, A., Chen, S. and Scali, C., Robust tuning of conventional controllers. Hydrocarbon Processing,
1990, November, 53-58.
49. Rivera, D.E., Morari, M. and Skogestad, S., Internal Model Control. 4. PID controller design. Industrial and
Engineering Chemistry Process Design and Development, 1986, 25(1), 252-265.
50. Chien, I.-L., IMC-PID controller design - an extension. Proceedings of the IFAC Adaptive Control of
Chemical Processes Conference, Copenhagen, Denmark, 1988, 147-152.
51. Thomasson, F.Y., Tuning guide for basic control loops. Proceedings of the 1997 process control, electrical
and information conference (TAAPI), 1997, 137-148.
52. Fruehauf, P.S., Chien, I.-L. and Lauritsen, M.D., Simplified IMC-PID tuning rules. Proceedings of the ISA/93
Advances in Instrumentation and Control Conference, McCormick Place, Chicago, Illinois, USA, 1993, 48,
1745-1766.
53. Chen, C.-L., Huang, H.-P. and Hsieh, C.-T. (1999). Tuning of PI/PID controllers based on specification of
maximum closed-loop amplitude ratio, Journal of Chemical Engineering of Japan, 32, 6, 783-788.
54. Ogawa, S., PI controller tuning for robust performance. Proceedings of the IEEE Conference on Control
Applications, 1995, 101-106.
55. Lee, Y., Park, S., Lee, M. and Brosilow, C., PID controller tuning for desired closed-loop responses for SI/SO
systems. AIChE Journal, 1998, 44(1), 106-115.
56. Isaksson, A.J. and Graebe, S.F., Analytical PID parameter expressions for higher order systems, Automatica,
1999, 35, 1121-1130.
57. Chun, D., Choi, J.Y. and Lee, J., Parallel compensation with a secondary measurement, Industrial Engineering
Chemistry Research, 1999, 38(4), 1575-1579.
58. McMillan, G.K., Control loop performance. Proceedings of the ISA/84 International Conference and
Exhibition. Advances in Instrumentation, Houston, Texas, USA, 1984, 39, 1, 589-603.
59. Shinskey, F.G., Feedback controllers for the process industries. McGraw-Hill Inc., New York, 1994.
60. Hwang, S.-H., Closed-loop automatic tuning of single-input-single-output systems. Industrial Engineering
Chemistry Research, 1995, 34(7), 2406-2417.
61. Hwang, S.-H. and Fang, S.-M., Closed-loop tuning method based on dominant pole placement. Chemical
Engineering Communications, 1995, 136, 45-66.
61a. Taguchi, H. and Araki, M., Two-degree-of-freedom PID controllers their functions and optimal tuning,
PID control), Terrassa, Spain, 2000, pp. 95-100.
62. Hwang, S.-H. and Chang, H.-C., A theoretical examination of closed-loop properties and tuning methods of
single-loop PI controllers. Chemical Engineering Science, 1987, 42(10), 2395-2415.
63. Pessen, D.W., A new look at PID-controller tuning. Transactions of the ASME. Journal of Dynamic Systems,
Measurement and Control, 1994, 116, 553-557.
64. Parr, E.A., Industrial Control Handbook, Vol. 3. BSP Professional Books, 1989.
65. Hang, C.C., Astrom, K.J. and Ho, W.K., Refinements of the Ziegler-Nichols tuning formula. IEE Proceedings,
Part D, 1991, 138(2), 111-118.
66. Hagglund, T. and Astrom, K.J., Industrial adaptive controllers based on frequency response techniques.
Automatica, 1991, 27(4), 599-609.
67. Leva, A., PID autotuning algorithm based on relay feedback. IEE Proceedings, Part D, 1993, 140(5), 328-338.
68. Astrom, K.J., Ziegler-Nichols auto-tuners. Report TFRT-3167, Department of Automatic Control, Lund
Institute of Technology, Lund, Sweden, 1982.
69. Calcev, G. and Gorez, R., Iterative techniques for PID controller tuning. Proceedings of the 34th Conference
on Decision and Control, New Orleans, LA ., USA, 1995, 3209-3210.
70. Cox, C.S., Arden, W.J.B. and Doonan, A.F., CAD Software facilities tuning of traditional and predictive
control strategies. Proceedings of the ISA/94 International Conference, Exhibition and Training Program.
Advances in Instrumentation and Control, Anaheim, CA., U.S.A., 1994, 49, 2, 241-250.
71. Vrancic, D., Peng, Y., Strmcnik, S. and Hanus, R., A new tuning method for PI controllers based on a process
step response. Proceedings of the CESA 96 IMACS Multiconference Symposium on Control, Optimisation
and Supervision, Lille, France, 1996, 2, 790-794.
72. Vrancic, D., Design of anti-windup and bumpless transfer protection. Part II: PID controller tuning by
multiple integration method. PhD thesis, University of Ljubljana, J. Stefan Institute, Ljubljana, Slovenia, 1996.
73. Vrancic, D., Peng, Y., Strmcnik, S. and Juricic, D., A multiple integration tuning method for filtered PID
controller. Proceedings of the IFAC 1999 14th World Congress, Beijing, China, 1999, Preprints, Paper 3b-02-3.
74. Chien, I.-L., Huang, H.-P. and Yang, J.-C., A simple multiloop tuning method for PID controllers with no
proportional kick, Industrial Engineering Chemistry Research, 1999, 38(4), 1456-1468.
75. Tyreus, B.D. and Luyben, W.L., Tuning PI controllers for integrator/dead time processes. Industrial
Engineering Chemistry Research, 1992, 31(11), 2625-2628.
76. Wang, L. and Cluett, W.R., Tuning PID controllers for integrating processes. IEE Proceedings - Control
Theory and Applications, 1997, 144(5), 385-392.
77. Rotach, V. Ya. (1995). Automatic tuning of PID-controllers expert and formal methods, Thermal
Engineering, 42, 10, 794-800.
78. Poulin, E. and Pomerleau, A.. PI settings for integrating processes based on ultimate cycle information. IEEE
79. Penner, A., Tuning rules for a PI controller. Proceedings of the ISA/88 International Conference and
Exhibition, Houston, Texas, USA, 1988, 1037-1051.
80. Srividya, R. and Chidambaram, M., On-line controllers tuning for integrator plus delay systems. Process
Control and Quality, 1997, 9, 59-66.
81. Tan, W., Liu, K. and Tam, P.K.S., PID tuning based on loop-shaping H control. IEE Proceedings - Control
82. Poulin, E. and Pomerleau. A., PID tuning for integrating and unstable processes. IEE Proceedings - Control
83. McAvoy, T.J. and Johnson, E.F. (1967). Quality of control problem for dead-time plants, Industrial and
Engineering Chemistry Process Design and Development, 6, 4, 440-446.
84. Lopez, A.M., Smith, C.L. and Murrill, P.W. (1969). An advanced tuning method, British Chemical
Engineering, 14, 11, 1553-1555.
85. Hougen, J.O., Measurement and Control Applications. Instrument Society of America, Research Triangle
Park, North Carolina, 1979.
86. De Paor, A.M. and O'Malley, M., Controllers of Ziegler-Nichols type for unstable processes with time delay.
International Journal of Control, 1989, 49(4), 1273-1284.
87. Venkatashankar, V. and Chidambaram, M., Design of P and PI controllers for unstable first order plus time
delay systems. International Journal of Control, 1994, 60(1), 137-144.
88. Chidambaram, M., Design of PI and PID controllers for an unstable first-order plus time delay system.
Hungarian Journal of Industrial Chemistry, 1995, 23, 123-127.
89. Rotstein, G.E. and Lewin, D.E., Simple PI and PID tuning for open-loop unstable systems. Industrial
Engineering Chemistry Research, 1991, 30(8), 1864-1869.
90. Ho, W.K. and Xu, W., PID tuning for unstable processes based on gain and phase-margin specifications.
IEE Proceedings - Control Theory and Applications, 1998, 145(5), 392-396.
91. Luyben, W.L., Tuning temperature controllers on openloop unstable reactors, Industrial Engineering
Chemistry Research, 1998, 37, 4322-4331.
91a. Hansen, P.D. (2000). Robust adaptive PID controller tuning for unmeasured load rejection, Preprints of the
Proceedings of PID 00: IFAC Workshop on digital control (Past, present and future of PID control),
Terrassa, Spain, pp. 487-494.
92. Poulin, E. and Pomerleau. A., Unified PID design method based on a maximum peak resonance specification.
IEE Proceedings - Control Theory and Applications, 1997, 144(6), 566-574.
93. Astrom, K.J. and Hagglund, T., Automatic tuning of PID Controllers, Instrument Society of America,
Research Triangle Park, North Carolina, 1988.
94. Sain, S.G. and Ozgen, C., Identification and tuning of processes with large deadtime. Control and Computers,
1992, 20(3), 73-78.
95. Cheng, G.S. and Hung, J.C., A Least-Squares Based Self-Tuning of PID Controller. Proceedings of the IEEE
South East Conference, Raleigh, North Carolina, USA, 1985, 325-332.
96. Gerry, J.P., How to control processes with large dead times, http://www.expertune.com/artdt.html, 1998.
97. Wang, F.-S., Juang, W.-S. and Chan, C.-T., Optimal tuning of PID controllers for single and cascade control
loops. Chemical Engineering Communications, 1995, 132, 15-34.
97a. Pi-Mira, J., Mateo, E., Sarrate-Estruch, R. and Quevedo-Casin, J., LS-3000 digital PID controller, Preprints of
the Proceedings of PID 00: IFAC Workshop on digital control (Past, present and future of PID control),
Terrassa, Spain, 2000, pp. 465-471.
98. Astrom, K.J. and Hagglund, T., Automatic tuning of simple regulators with specifications on phase and
amplitude margins. Automatica, 1984, 20(5), 645- 651.
99. Li, Z., Su., X. and Lin, P., A practical algorithm for PID auto-tuning. Advances in Modelling and Analysis C,
ASME Press, 1994, 40(2), 17-27.
100. Suyama, K., A simple design method for sampled-data PID control systems with adequate step responses.
Proceedings of the International Conference on Industrial Electronics, Control, Instrumentation and
Automation, 1992, 1117-1122.
101. Juang, W.-S. and Wang, F.-S., Design of PID controller by concept of Dahlins Law. Journal of the Chinese
Institute of Chemical Engineers, 1995, 26(2), 133-136.
102. Camacho, O.E., Smith, C. and Chacon, E., Toward an implementation of sliding mode control to chemical
processes, Proceedings of the IEEE International Symposium on Industrial Electronics, 1997, 3, 1101-1105.
103. Ho, W.K., Lim, K.W. and Xu, W., Optimal gain and phase margin tuning for PID controllers. Automatica,
1998, 34(8), 1009-1014.
104. Ho, W.K., Lim, K.W., Hang, C.C. and Ni, L.Y., Getting more phase margin and performance out of PID
controllers, Automatica, 1999, 35, 1579-1585.
104a. Morilla, F., Gonzalez, A. and Duro, N., Auto-tuning PID controllers in terms of relative damping, Preprints of
the Proceedings of PID 00: IFAC Workshop on digital control (Past, present and future of PID control),
105. Morari, M. and Zafiriou, E., Robust process control. Prentice-Hall Inc., Englewood Cliffs, New Jersey, 1989.
106. Horn, I.G., Arulandu, J.R., Gombas, C.J., VanAntwerp, J.G. and Braatz, R.D., Improved filter design in internal
model control. Industrial Engineering Chemistry Research, 1996, 35(10), 3437-3441.
106a. Normey-Rico, J.E., Alcala, I., Gomez-Ortega, J. and Camacho, E.F. (2000). Robust PID tuning application to a
mobile robot path tracking problem, Preprints of the Proceedings of PID 00: IFAC Workshop on digital
control (Past, present and future of PID control), Terrassa, Spain, pp. 648-653.
107. Witt, S.D. and Waggoner, R.C., Tuning parameters for non-PID three-mode controllers. Hydrocarbon
Processing, 1990, June, 74-78.
108. Kaya, A. and Scheib, T.J., Tuning of PID controls of different structures. Control Engineering, 1988, July, 6265.
109. Tsang, K.M. and Rad, A. B., A new approach to auto-tuning of PID controllers. International Journal of
Systems Science, 1995, 26(3), 639-658.
110. Tsang, K.M., Rad, A.B. and To, F.W., Online tuning of PID controllers using delayed state variable filters.
Proceedings of the IEEE Region 10 Conference on Computer, Communication, Control and Power
Engineering, 1993, 4, 415-419.
111. Hang, C.C. and Astrom, K.J., Refinements of the Ziegler-Nichols tuning formuale for PID auto-tuners.
Proceedings of the ISA/88 International Conference and Exhibition. Advances in Instrumentation, 1988, 43,
3, 1021-1030.
112. Hang, C.-C. and Cao, L., Improvement of transient response by means of variable set-point weighting. IEEE
Transactions on Industrial Electronics, 1996, 43(4), 477-484.
113. Gong, X., Gao, J. and Zhou, C., Extension of IMC tuning to improve controller performance. Proceedings of
the IEEE International Conference on Systems, Man and Cybernetics, 1996, 1770-1775.
114. VanDoren, V.J., Ziegler-Nichols methods facilitate loop tuning. Control Engineering, 1998, December.
114a. Argelaguet, R., Pons, M., Quevedo, J. and Aguilar, J., A new tuning of PID controllers based on LQR
optimization, Preprints of the Proceedings of PID 00: IFAC Workshop on digital control (Past, present and
future of PID control), Terrassa, Spain, 2000, pp. 303-308.
115. Blickley, G.J., Modern control started with Ziegler-Nichols tuning. Control Engineering, 1990, 2 October, 1117.
116. De Paor, A.M., A fiftieth anniversary celebration of the Ziegler-Nichols PID controller. International Journal
of Electrical Engineering Education, 1993, 30, 303-316.
117. Corripio, A.B., Tuning of industrial control systems. Instrument Society of America, Research Triangle Park,
North Carolina, 1990.
118. Mantz, R.J. and Tacconi, E.J., Complementary rules to Ziegler and Nichols' rules for a regulating and tracking
controller. International Journal of Control, 1989, 49(5), 1465-1471.
119. Atkinson, P. and Davey, R.L., A theoretical approach to the tuning of pneumatic three-term controllers.
Control, 1968, March, 238-242.
120. Perry, R.H. and Chilton, C.H., Chemical engineers handbook, McGraw-Hill, New York, 5th edition, 1973.
121. Luo, K.-N., Kuo, C.-Y. and Sheu, L.-T. (1996). A novel method for fuzzy self-tuning PID controllers,
Proceedings of the Asian Fuzzy Systems Symposium, 194-199.
122. Yu, C.-C. (1999). Autotuning of PID controllers, Advances in Industrial Control Series, Springer-Verlag
London Ltd.
123. Karaboga, D. and Kalinli, A., Tuning PID controller parameters using Tabu search algorithm. Proceedings of
the IEEE International Conference on Systems, Man and Cybernetics, 1996, 134-136.
124. Hang, C.C. and Astrom, K.J., Practical aspects of PID auto-tuners based on relay feedback. Proceedings of
the IFAC Adaptive control of Chemical Processes Conference, Copenhagen, Denmark, 1988, 153-158.
125. Shin, C.-H., Yoon, M.-H. and Park, I.-S., Automatic tuning algorithm of the PID controller using two Nyquist
points identification. Proceedings of the Society of Instrument and Control Engineers annual conference,
Tokyo, Japan, 1997, 1225-1228.
126. Harriott, P., Process Control, McGraw-Hill, New York, 1964.
127. Zhang, G., Shao, C. and Chai, T., A new method for independently tuning PID parameters. Proceedings of
the 35th Conference on Decision and Control, Kobe, Japan, 1996, 2527-2532.
127a. Garcia, R.F. and Castelo, F.J.P., A complement to autotuning methods on PID controllers, Preprints of the
128. Fu, M., Olbrot, A.W. and Polis, M.P., Comments on 'Optimal gain for proportional-integral-derivative
feedback'. IEEE Control Systems Magazine, 1989, January, 100-101.
129. Pessen, D.W., How to tune in a three-mode controller. Instrumentation, 1954, 7(3), 29-32.
130. Grabbe, E.M., Ramo, S. and Woolridge, D.E. (Editors), Handbook of automation, computation and control.
Vol 3: Systems and Components, John Wiley and Sons, New York, 1961.
131. Pessen, D.W., Optimum three-mode controller settings for automatic start-up. Transactions of the ASME,
1953, July, 843-849.
132. Ford, R.L., The determination of the optimum process-controller settings and their confirmation by means of
an electronic simulator. Proceedings of the IEE, Part 2, 1953, 101(80), April, 141-155, 173-177.
133. Luyben, W.L., Tuning proportional-integral-derivative controllers for integrator/deadtime processes.
Industrial Engineering Chemistry Research, 1996, 35(10), 3480-3483.
134. Belanger, P.W. and Luyben, W.L., Design of low-frequency compensators for improvement of plantwide
regulatory performances. Industrial Engineering Chemistry Research, 1997, 36(12), 5339-5347.
134a. Pecharroman, R.R. and Pagola, F.L., Control design for PID controllers auto-tuning based on improved
identification, Preprints of the Proceedings of PID 00: IFAC Workshop on digital control (Past, present
and future of PID control), Terrassa, Spain, 2000, pp. 89-94.
134b. Pecharroman, R.R., Private communication, 9 May 2000.
135. Zhang, W., Xu, X. and Sun, Y., Quantitative performance design for integrating processes with time delay,
Automatica, 1999, 35, pp. 719-723.
136. Tan, W., Liu, J. and Sun, W., PID tuning for integrating processes. Proceedings of the 1998 IEEE
International Conference on Control Applications, Trieste, Italy, 1998, 2, 873-876.
137. Chien, I.-L. and Fruehauf, P.S., Consider IMC tuning to improve controller performance. Chemical
Engineering Progress, 1990, October, 33-41.
138. Oubrahim, R. and Leonard, F., PID tuning by a composed structure. Proceedings of the UKACC International
Conference on Control 98, Swansea, Wales, 1998, 2, 1333-1338.
139. Sung, S.W., O, J., Lee, I.-B., Lee, J. and Yi, S.-H., Automatic tuning of PID controller using second-order plus
time delay model. Journal of Chemical Engineering of Japan, 1996, 29(6), 991-999.
140. Ho, W.K., Hang, C.C. and Cao, L.S., Tuning of PID controllers based on gain and phase margin
specifications. Automatica, 1995, 31(3), 497-502.
141. Ho, W.K., Hang, C.C. and Zhou, J., Self-tuning PID control of a plant with under-damped response with
specifications on gain and phase margins. IEEE Transactions on Control Systems Technology, 1997, 5(4),
446-452.
142. Ho, W.K., Hang, C.C., Zhou, J.H. and Yip, C.K., Adaptive PID control of a process with underdamped
response. Proceedings of the Asian Control Conference, Tokyo, Japan, 1994, 335-338.
143. Wang, Q.-G., Lee, T.-H., Fung, H.-W., Bi, Q. and Zhang, Y., PID tuning for improved performance. IEEE
144. Wang, Y.-G. and Shao, H.-H. (1999). PID autotuner based on gain- and phase-margin specification,
Industrial Engineering Chemistry Research, 38, 3007-3012.
145. Pemberton, T.J., PID: The logical control algorithm II. Control Engineering, 1972, 19(7), 61-63.
146. Smith, C.L., Corripio, A.B. and Martin, J., Controller tuning from simple process models. Instrumentation
Technology, 1975, December, 39-44.
147. Wang, T.-S. and Clements, W.C., Adaptive multivariable PID control of a distillation column with unknown
and varying dead time. Chemical Engineering Communications, 1995, 132, 1-13.
147a. Gorez, R. and Klan, P., Nonmodel-based explicit design relations for PID controllers, Preprints of the
147b. Seki, H., Ogawa, M. and Ohshima, M. (2000). Retuning PID temperature controller for an unstable gasphase polyolefin reactor, Preprints of the Proceedings of PID 00: IFAC Workshop on digital control
(Past, present and future of PID control), Terrassa, Spain, pp. 473-478.
148. Landau, I.D. and Voda, A., An analytical method for the auto-calibration of PID controllers. Proceedings of
the 31st Conference on Decision and Control, Tucson, Arizona, USA, 1992, 3237-3242.
149. Jahanmiri, A. and Fallahi, H.R., New methods for process identification and design of feedback controller.
Transactions of the Institute of Chemical Engineers, 1997, 75(A), July, 519-522.
150. Hansen, P.D., Controller structure and tuning for unmeasured load disturbance. Proceedings of the American
Control Conference, Philadelphia, Pennsylvania, 1998, 1, 131-136.
151. Hang, C.C., Ho, W.H. and Cao, L.S., A comparison of two design methods for PID controllers. ISA
Transactions, 1994, 33, 147-151.
152. Polonyi, M.J.G., PID controller tuning using standard form optimisation. Control Engineering, 1989, March,
102-106.
153. Valentine, C.C. and Chidambaram, M., PID control of unstable time delay systems. Chemical Engineering
Communications, 1997, 162, 63-74.
154. Huang, C.-T. and Lin, Y.-S., Tuning PID controller for open-loop unstable processes with time delay.
Chemical Engineering Communications, 1995, 133, 11-30.
155. Skoczowski, S. and Tarasiejski, L., Tuning of PID controllers based on gain and phase margin specifications
using Strejcs process model with time delay. Proceedings of the Third International Symposium on
Methods and Models in Automation and Robotics (MMAR 96), Miedzyzdroje, Poland, 1996, 765-770.
156. Lennartson, B. and Kristiansson, B., Pass band and high frequency robustness for PID control. Proceedings
of the 36th IEEE Conference on Decision and Control, San Diego, California, U.S.A., 1997, 2666-2671.
157. Kristiansson, B. and Lennartson, B., Robust design of PID controllers including auto-tuning rules.
Proceedings of the American Control Conference, 1998, 5, 3131-3132.
158. Kristiansson, B. and Lennartson, B., Optimal PID controllers for unstable and resonant plants. Proceedings
of the Conference on Decision and Control, 1998.
158a. Kristiansson, B. and Lennartson, B., Near optimal tuning rules for PI and PID controllers, Preprints of the
159. Vrancic, D., Peng, Y. and Strmcnik, S., A new PID controller tuning method based on multiple integrations.
Control Engineering Practice, 1999, 7(5), 623-633.
160. Hwang, S.-H. (1995). Closed-loop automatic tuning of single-input-single-output systems, Industrial
Engineering Chemistry Research, 34, 2406-2417.
161. Ferretti, G., Maffezzoni, C. and Scattolini, R. (1991). Recursive estimation of time delay in sampled systems,
Automatica, 27, 653-661.
162. Nishikawa, Y., Sannomiya, N., Ohta, T. and Tanaka, H. (1984). A method for auto-tuning of PID control
parameters, Automatica, 20, 321-332.
163. Lee, J. and Sung, S.W. (1993). Comparison of two identification methods for PID controller tuning, AIChE
Journal, 39, 695-697.
164. Deshpande, P.B. (1980). Process identification of open-loop unstable systems, AIChE Journal, 26, 305-308.
165. Pecharroman, R.R. and Pagola, F.L. (1999). Improved identification for PID controllers auto-tuning,
Proceedings of the 5 th European Control Conference (ECC 99), Karlsruhe, Germany, Paper F453, BA-12.
Appendix 1: List of symbols used (more than once) in the paper.

a 1, a2 = PID filter parameters
A m = gain margin
b = setpoint weighting factor
b1 = PID filter parameter

c = derivative term weighting factor
E(s) = Desired variable, R(s), minus controlled variable, Y(s)
FOLPD model = First Order Lag Plus time Delay model
FOLIPD model = First Order Lag plus Integral Plus time Delay model
G c ( s) = PID controller transfer function

G p ( j ) = process transfer function at frequency
G p ( j ) = magnitude of G p ( j ) ,
Gp ( j) = phase of G p ( j )
IAE = integral of absolute error

IMC = internal model controller
IPD model = Integral Plus time Delay model
ISE = integral of squared error
ISTES = integral of squared time multiplied by error, all to be squared
ISTSE = integral of squared time multiplied by squared error
ITAE = integral of time multiplied by absolute error
ki = integral gain of the parallel PID controller

k p = proportional gain of the parallel PID controller
kd = derivative gain of the parallel PID controller

Kc = Proportional gain of the controller
Km = Gain of the process model
Ku = Ultimate gain
K25% = proportional gain required to achieve a quarter decay ratio
m = multiplication parameter in the lead controller
Ms = closed loop sensitivity

N = determination of the amount of filtering on the derivative term
PP = perturbance peak = peak of system output when unit step disturbance is added
R(s) = Desired variable
RT= recovery time = time for perturbed system output (when a unit step disturbance is added) to come to its final
value
SOSPD model = Second Order System Plus time Delay model
Td = Derivative time of the controller

TCL = desired closed loop system time constant
Tf = Time constant of the filter in series with the PID controller
Ti = Integral time of the controller
Tm = Time constant of the FOLPD process model
Tm1, Tm2 ,Tm3 = Time constants of the higher order process models
Tu = Ultimate time constant
T25% = period of the quarter decay ratio waveform
TOLPD model = Third Order Lag Plus time Delay model
TS = settling time
U(s) = manipulated variable
Y(s) = controlled variable
= Parameter that determines robustness of compensated system.
m = damping factor of an underdamped process model, = damping factor of the compensated system
= 1 K m Ku
= phase lag,
m = phase margin,
= phase lag at an angular frequency of
c = phase corresponding to the crossover frequency

m = time delay of the process model,
= angular frequency,
= m ( m + Tm )
u = ultimate frequency,
= angular frequency at a phase lag of

O.dyer PID

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

O.dyer PID

Uploaded by

Copyright:

Available Formats

PI and PID controller tuning rules for time delay processes: a summary

Technical Report AOD-00-01, Edition 1

m = t 67% 1.1Tm . K m assumed known (Chen and Yang [23a]).

C. Integral plus time delay (IPD) model ( G m ( s) =

* Method 1: Method of moments (Astrom and Hagglund [3]).

Second order system plus time delay (SOSPD) model ( G m (s) =

Method 1: Km , Tm1 , Tm2 , m or Km , Tm1 , m , m assumed known.

Method 1: Km , Tm , m assumed known.

H. Third order system plus time delay (TOLPD) model ( G m (s) =

Method 1: Km , Tm1 , Tm2 , Tm3 , m known.

I. Unstable first order lag plus time delay model ( G m (s) =

J. Unstable second order system plus time delay model ( G m (s) =

Method 1: Km , Tm1 , Tm2 , m known.

(1 + sTm1 )(1 + sTm2 )

Method 1: Km , Tm1 , Tm2 , Tm3 , m known.

(1 + sTm1 )(1 + sTm 2 )

Method 1: Km , Tm1 , Tm2 , Tm3 , m known.

K m (1 + b1s + b2 s2 + b3s3 + b4 s4 + b ss5 ) e s

* Method 1: Strejcs method

minimisation tuning rule is defined by Taguchi and Araki [61a].

[3] page 215.

Tables 8-11: PI tuning rules IPD model G m ( s) =

[3] pages 210-212.

minimisation tuning rule is defined by Taguchi and Araki [61a].

Table 20: PI tuning rules - unstable FOLPD model

Table 21: PI tuning rules - unstable SOSPD model

(a) Direct synthesis: Hansen [91a].

Tables 23-40: PID tuning rules - FOLPD model

is defined by Astrom and Hagglund [3] pages 208-210.

by Normey-Rico et al. [106a].

Table 28: Classical controller G c ( s) = K c 1 +

(a) ******: Astrom and Hagglund [3] page 246.

synthesis tuning rule is defined by Tsang et al. [110].

et al. [74] and Argelaguet et al. [114a].

defined by Mantz and Tacconi [118].

rule is defined by Astrom and Hagglund [3] page 217.

defined by VanDoren [114].

has been defined by Normey-Rico et al. [106a].

Table 55: I-PD controller U(s) =

Chien et al. [74].

Tables 58-67: PID tuning rules - FOLIPD model

has been defined by Astrom and Hagglund [3] - pages 212-213.

Table 65: Ideal controller with set-point weighting:

. One direct synthesis tuning rule has

Table 67: Two degree of freedom controller:

Tables 68-79: PID tuning rules - SOSPD model

. One ***** tuning rule has been

defined by Hougen [85].

defined. The references are:

has been defined by Oubrahim and Leonard [138].

Table 80: Controller U(s) = Kc (1 +

(1 + sTm1 )(1 + sTm2 )

(1 + sTm1 )(1 + sTm 2 )

defined by Wang et al. [97].

Table 89-90: PID tuning rules - TOLPD model

defined by Polonyi [153].

Tables 91-93: PID tuning rules - unstable FOLPD model

(a) Ultimate cycle: McMillan [58]

defined; the references are

Skoczowski and Tarasiejski [156]

and Klan [147a].

** some more information needed.

Table A: Model structure and tuning rules a summary for PI controllers