You are on page 1of 14

ISA Transactions 39 (2000) 205218

www.elsevier.com/locate/isatrans

Sliding mode control: an approach to regulate nonlinear


chemical processes
Oscar Camacho a,*, Carlos A. Smith b
a
Departamento de Circuitos y Medias, Universidad de Los Andes, Merida 5101, Venezuela
b
Chemical Engineering Department, University of South Florida, Tampa, Florida, USA

Abstract
A new approach for the design of sliding mode controllers based on a rst-order-plus-deadtime model of the process,
is developed. This approach results in a xed structure controller with a set of tuning equations as a function of the
characteristic parameters of the model. The controller performance is judged by simulations on two nonlinear chemical
processes # 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Sliding mode control; Variable structure control; Nonlinear chemical processes

1. Introduction procedures of SMC presents disadvantages in


their application to chemical processes.
Sliding mode control (SMC) is a robust and An ecient alternative modeling method for
simple procedure to synthesize controllers for lin- process control is the use of empirical models,
ear and nonlinear processes. To develop a sliding which use low order linear models with deadtime.
mode controller, SMCr, knowledge of the process Most times, rst-order-plus deadtime (FOPDT)
model relating the controlled variable, Xc t, to models are adequate for process control analysis
the manipulated variable, Ut, is necessary. and design. But, these reduced order models pre-
However, there are two problems with the use of a sent uncertainties arising from imperfect knowl-
model as far as chemical processes are concerned. edge of the model, and the process nonlinear
First, the development of a complete model is dif- eects contribute to performance degradation of
cult due mainly to the complexity of the process the controllers.
itself, and to the lack of knowledge of some process Conventional controllers, such as PID, lead-lag
parameters. Second, most process models relating or Smith predictors, are sometimes not suciently
the controlled and the manipulated variables are versatile to compensate for these eects. Thus, a
of higher-order. Generally, the SMC procedure SMCr could be designed to control nonlinear sys-
produces a complex controller, which could con- tems with the assumption that the robustness of
tain four or more parameters resulting in a dicult the controller will compensate for modeling errors
tuning job. Therefore, the use of the traditional arising from the linearization of the nonlinear
model of the process.
* Corresponding author. Tel.: +58-74-402-891; fax: +58- The aim of this paper is to design a SMCr based
74-402-890. on a rst-order-plus-deadtime (FOPDT) model of
E-mail address: ocamacho@ing.ula.ve (O. Camacho). the actual process. The overall idea is to develop a
0019-0578/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0019-0578(99)00043-9
206 O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218

general SMCr, which can be used for self-regulat- is simple to perform, and provides adequate models
ing chemical processes. The parameters of the for many applications. The curve is obtained by
model, process gain, K, process time constant, , introducing a step change in the output from the
and process deadtime, t0 , are used to obtain the controller and recording the transmitter output.
initial estimates of the tuning terms in the SMCr. From the process curve shown in the gure, and
This article is organized as follows. Section 2 the procedure presented in the reference, the
briey presents the process model. Section 3, pre- numerical values of the terms in the FOPDT
sents some basic concepts of the SMC method. model given in Eq. 1 are obtained
Section 4 shows the procedure to design a SMCr
using the FOPDT model. Tuning equations for Xs Ket0 s
1
the controller are also given in this section. In Us s 1
Section 5 the simulation of the SMCr for two
nonlinear chemical processes is presented. Section where Xs is the Laplace transform of the con-
6 concludes the paper. trolled variable, the transmitter output, and Us
is the Laplace transform of the manipulated vari-
able, the controller output. Both Xs and Us are
2. Process model deviation variables. In this paper we use the unit
of Xs as fraction of the transmitter output, frac-
The process reaction curve, Fig. 1, is an often- tion TO; the unit of Us is fraction of the con-
used method for identifying dynamic models 1. It troller output, fraction CO. K, t0 and  were
previously dened.

3. Basic concepts about sliding mode control

Sliding Mode Control is a technique derived


from variable structure control (VSC) which was
originally studied by [2]. The controller designed
using the SMC method is particularly appealing
due to its ability to deal with nonlinear systems
and time-varying systems [35]. The robustness to
the uncertainties becomes an important aspect in
designing any control system.
The idea behind SMC is to dene a surface
along which the process can slide to its desired
nal value; Fig. 2 depicts the SMC objective. The
structure of the controller is intentionally altered
as its state crosses the surface in accordance with a
prescribed control law. Thus, the rst step in SMC
is to dene the sliding surface St. St is chosen
to represent a desired global behavior, for instance
stability and tracking performance; The St
selected in this work, presented by [4], is an inte-
gral-dierential equation acting on the tracking-
error expression.
 n t
d
St l etdt 2
Fig. 1. Process reaction curve. dt 0
O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218 207

where fXt; Rt is a function of the controlled


variable, and the reference value.
The discontinuous part, UD t, incorporates a
nonlinear element that includes the switching ele-
ment of the control law. This part of the controller
is discontinuous across the sliding surface.

St
UD t KD 6
St 

where KD is the tuning parameter responsible for


the reaching mode.  is a tuning parameter used to
Fig. 2. Graphical interpretation of SMC.
reduce the chattering problem. Chattering is a
high-frequency oscillation around the desired
where et is the tracking error, that is, the dier- equilibrium point. It is undesirable in practice,
ence between the reference value or set point, Rt, because it involves high control activity and also
and the output measurement, Xt, or et can excite high-frequency dynamics ignored in the
Rt Xt. l is a tuning parameter, which helps modeling of the system [3,4,6].
to dene St; This term is selected by the In summary, the control law usually results in a
designer, and determines the performance of the fast motion to bring the state onto the sliding
system on the sliding surface, n is the system surface, and a slower motion to proceed until a
order. desired state is reached.
The objective of control is to ensure that the
controlled variable be equal to its reference value
at all times, meaning that et and its derivatives 4. SMCr synthesis from an FOPDT model of
must be zero. Once the reference value is reached, the process
Eq. (2) indicates that St reaches a constant
value. To maintain St at this constant value, This section presents the development of a
meaning that et is zero at all times; it is desired general SMCr, for self-regulating processes, using
to make a rst-order-plus-deadtime (FOPDT) process
model. The FOPDT model is an approximation to
dSt the actual higher-order model. The development
0 3
dt of this controller signicantly simplies the appli-
cation of sliding mode control theory to chemical
Once the sliding surface has been selected, processes.
attention must be turned to design of the control The literature reviewed does not reveal a simple
law that drives the controlled variable to its refer- and practical method to apply SMC to process
ence value and satises Eq. (3). The SMC control with dead time [79]. In this chapter, a SMCr
law, Ut, consists of two additive parts; a con- structure based on the FOPDT model of the
tinuous part, UC t, and a discontinuous part, actual process is designed. Thus, the rst step is to
UD t, [6]. That is propose a way to handle the deadtime term.
The deadtime can be approximated in two dif-
Ut UC t UD t 4 ferent ways. A rst-order Taylor series approx-
imation to the deadtime term produces.
The continuous part is given by
1
et0 s 7
UC t fXt; Rt 5 t0 s 1
208 O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218

The above approximation can also be written as a Substituting Eq. (7) into Eq. (1) produces
rst-order Pade approximation
Xs K
1 0:5t0 s 9
et0 s 8 Us s 1t0 s 1
1 0:5t0 s
Fig. 3 shows a comparison among the deadtime In dierential equation form
term and the rst-order Taylor series and Pade
approximations. The gure shows that the Pade d2 Xt dXt
t0  t0  Xt KUt 10
approximation works very well between 0 and 1 but dt2 dt
beyond the approximation brakes down. On other
hand, the Taylor series approximation improves and since this is a second-order dierential equa-
as t0 increases. tion, n 2, from Eq. (2) St becomes
In [10] is shown that the rst-order Taylor
approximation or the Pade approximation can be t
det
considered as good approximations for the dead- St l1 et l0 etdt 11
dt 0
time term for chemical processes.
The next section shows the development of a
SMCr using both approximations. Where l1 2l and l0 l2
From Eq. (3)
4.1. SMCr development based on a rst-order
Taylor series approximation dSt d2 et det
2
l1 l0 e t 0 12
dt dt dt
In this section a SMCr is developed based on the
rst-order Taylor series expansion. Additionally, a Substituting the denition of the error,
rule to choose the tuning parameters will also be et Rt Xt, into the rst two terms of the
presented. above equation gives

Fig. 3. Comparison among ex (1), Taylor (2) and Pade (3) approximations.
O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218 209
 
d2 Rt d2 Xt dRt dXt with
l 1
dt2 dt2 dt dt
St signK
l0 et 0 13
 t  18b
dXt
l1 et l0 etdt
Solving for the highest derivative from Eq. (10), dt 0
substituting it into the Eq. (13), and solving for
Ut provides the continuous part of the controller Eqs. (18a) and (18b) constitute the controller
" equations to be used. These equations present
t   t   dXt Xt advantages from process control point of view,
0 0
UC t l1
K t0  dt t0  rst they have a xed structure depending on the
# l's parameters and the characteristic parameters
d2 Rt dRt of the FOPDT model, and second the action of
l0 et l1 14
dt2 dt the controller is considered in the sliding surface
equation, by including the term signK, in Eq.
This procedure, involving Eqs. (11) and (13), to 18b. Note, that signK only depends on the static
obtain the expression for the continuous part of gain, therefore it never switches. From an indus-
the controller is known in the SMC theory as the trial application perspective, Eq. (18b) represents a
equivalent control procedure [2]. PID algorithm [12].
In [11] is shown that the derivatives of the To complete the SMCr, it is necessary to have a
reference value can be discarded, without any set of tuning equations. For the tuning equations
eect on the control performance, resulting in a as rst estimates, using the NelderMead search-
simpler controller. Thus, ing algorithm [13], the following equations were
" obtained [11].
t   t  
dXt
0 0
UC t l1 . For the continuous part of the controller and
K t0  dt
# the sliding surface
Xt
l0 et 15 t0 
t0  l1 time1 19a
t0 

UC t can be simplied by letting,  


1 t0  2
l0 time2 19b
t0  4 t0 
l1 16
t0 
. For the discontinuous part of the controller

It has been shown that this choice for l1 is the  


0:51  0:76
best for the continuous part of the controller [11]. KD fraction CO 19c
jKj t0
To assure that the sliding surfaces behave as a
critical or overdamped system, l0 should be  0:68 0:12jKjKD l1
19d
l21 fraction TO=time
l0 4 17
4
Eqs. (19c) and (19d) are used when the signals
Then, the complete SMCr can be represented as from the transmitter and controller are in frac-
follows tions (01). Sometimes, the control systems work
t  Xt  in percentages that is, the signals are in% (0100)
0 St of range. In these cases the values of KD and  are
Ut l0 et KD 18a
K t0  St  multiplied by 100.
210 O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218

4.2. SMCr development based on the Pade Thus, from the previous results, the controller
approximation equation to be used is that obtained from the
Taylor series approximation. The next part illus-
This section contains the development of the trates the controller performance.
control law when the deadtime term of the FOPDT
process model is approximated by the Pade
approximation, Eq. (8). The procedure followed in 5. Simulation results
this section is similar to that one presented in the
previous part. Substituting Eq. (8) into Eq. (1), This section simulates the control performance
gives of the SMCr designed and given in Eqs. (18a) and
(18b). The rst process, a mixing tank, compares
XC s K2 t0 s the performance of the SMCr with respect to a
20 PID controller. The second process, a chemical
Us s 12 t0 s
reactor, presents further performance character-
istics.
Using a similar procedure as shown above, the
continuous part of the controller, UC s, is 5.1. Mixing tank

2 "  3 Consider the mixing tank shown in Fig. 4. The


2
 2 t0
6 s l1 s Rs l1 s 7 tank receives two streams, a hot stream, W1 t,
6 t0  7 and a cold stream, W2 t. The outlet temperature
6 # 7
6 7 is measured at a point 125 ft downstream from the
6 2 7
6 XC s l0 es 7 tank. The following assumptions are accepted
6
6 t0  7
7
UC s 6 7
K6 2 7 . The liquid volume in the tank is considered
6 s 7
6 t0 7 constant
6 7
6 7 . The tank contents are well mixed
6 7
4 5 . The tank and the pipe are well insulated.

21 The temperature transmitter is calibrated for a


range of 100 to 200 F. Table 1 shows the steady-
state conditions and other operating information.
Eq. (21) has a pole 2=t0 on the right side of The following equations constitute the process
the complex plane. Thus, the continuous part of model
the controller contains an unstable term.
Eq. (20) represents a nonminimun phase system.
Hence, the equivalent control procedure applied
directly over this kind of systems produce unstable
controllers. An approach to solve the previous
problem, and that permit the use of SMC to non-
minimun phase processes is presented in [10].
In summary, up to now, the synthesis of a
SMCr has been shown from the linearization of a
nonlinear chemical process. The linear model
representing the nonlinear chemical process is an
FOPDT model. The characteristic parameters of
the FOPDT model also are used in the tuning
equations. Fig. 4. Mixing tank.
O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218 211

Table 1 where
Design parameters and steady-state values

Variable Value Variable Value W1 t = mass ow of hot stream, lb/min


W2 t = mass ow of cold stream, lb/min
W1 250.00 lb/min V 15 ft3 Cp = liquid heat capacity at constant pressure,
W2 191.17 lb/min TO 0.5
Cp1 0.8 Btu/lb- F Vp 0.478
Btu/lb- F
Cp2 1.0 Btu/lb- F CVL 12 gpm/psi1/2 Cv = liquid heat capacity at constant volume,
Cp3 ; Cv3 0.9 Btu/lb- F Pv 16 psi Btu/lb- F
Set point 150 F T 0.5 min T1 t = hot ow temperature,  F
T1 250 F vp 0.4 min T2 t = cold ow temperature,  F
T2 50 F A 0.2006 ft2
T3 150 F L 125 ft
T3 t = liquid temperature in the mixing tank,  F
 62.4 lb/ft3 m 0.478 CO T4 t = equal to T3 t delayed by t0 ,  F
t0 = deadtime or transportation lag, min
 = density of the mixing tank contents,
lbm/ft3
. Energy balance around mixing tank
V = liquid volume, ft3
W1 tCp1 tT1 t W2 tCp2 tT2 t TOt = transmitter output signal on a scale from
0 to 1
W1 t W2 tCp3 tT3 t VP t = valve position, from 0 (valve closed) to
1 (valve open)
dT3 t mt = fraction of controller output, from 0 to 1
VCv3 22 CVL = valve ow coecient, gpm/psi1/2
dt
Gf = specic gravity, dimensionless
. Pipe delay between the tank and the sensor Pv = pressure drop across the valve, psi
location T = time constant of the temperature sensor,
min
T4 t T3 t t0 23
Vp = time constant of the actuator, min
. Transportation lag or delay time A = pipe cross section, ft2
L = pipe length, ft
LA
t0 24
W1 t W2 t Following the procedure, presented in Section 2,
to obtain the parameters of the FOPDT model
. Temperature transmitter yields: K 0:78 fraction TO/fraction CO, 
  2:32 min., and t0 2:97 min. Using these values
dTOt 1 T4 t 100
TOt 25 the tuning parameters for the SMCr are
dt T 100

. Valve position l1 0:76 min1 ; KD 0:54 fraction CO

dVp t 1   l0 0:147 min2 ;  0:719 fraction TO=min


mt Vp t 26
dt Vp
The tuning parameters for the PI controller are
. Valve equation KC 0:5 and 1 2:32 min, using the tuning
formulas for Dahlin synthesis, which produce
500 p
W2 CVL Vp t Gf Pv 27 smoother responses than ZieglerNichols tuning
60 equations, working better for process with dead-
. Sliding mode controller (SMCr) time [1]. Note that the comparison is done using
the initial tuning parameters for both controllers,
Ut UC t UD t 28 to show the good performance obtained for the
212 O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218

SMCr initial tuning equations, but they can be mance of the PID controller. To recover stability,
adjusted, ne tuning, until acceptable control per- new tunings are required for the PI controller
formance be obtained. while none are required for the SMCr.
Please note that the controller equations, Eqs. In spite of the controller being synthesized using
(18a) and (18b), were developed using deviation a Taylor approximation and the tuning equations,
variables. The following changes the ``deviation Eqs. 19a19d, are empirical, the proposed method
variables'' in the controller to ``actual variables'' can be successfully used in processes with a dead-
time to time constant ratio larger than one. In our
Ut mt m experience, they can be applied for t0 = around 3.

and

Xt TOt TO

et Rt TOt

where mt is the controller output, in fraction


CO, TOt is the transmitter output, in fraction,
and Rt is the reference value, or set point, frac-
tion TO. The overbars indicate steady-state
values.
Since the process gain is negative, signK is
negative, the controller equation to be used is
 
t0  1 
mt m TOt TO l0 et
K t0 
KD St
18c
St 

with

t
dTOt
St l1 et l0 etdt 18d
dt 0

Fig. 5 shows the response of the temperature,


T4 t, when the ow of hot water changes from
250 lb/min to 200 lb/min, then to 175 lb/min, to
150 lb/min, and nally to 125 lb/min. The curves
clearly show that as the operating conditions
change, the performance of the PID controller
degrades, while that of the SMCr maintains its
performance and stability. In this case, as the ow
of hot water decreases, with a corresponding
decrease in cold water, the deadtime between the
tank and the sensor increases. This increase in
deadtime certainly adversely aects the perfor- Fig. 5. Temperature response under SMCr and PID controller.
O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218 213

5.2. Chemical reactor . Energy balance on reactor contents

dTt Ft HR
The reactor shown in Fig. 6 is a continuous Tit Tt kC2A
stirred tank where the exothermic reaction A!B dt V Cp
takes place. To remove the heat of reaction the UA
Tt TC t 30
reactor is surrounded by a jacket through which a V Cp
cooling liquid ows.
The following assumptions are accepted . Energy balance on jacket

. heat losses from the jacket to the surround- dTC t UA


Tt TC t
ings are negligible dt VC C Cpc
. densities and heat capacities of the reactants FC t
and products are both equal and constant TC t Tci t 31
VC
. the heat of reaction is constant.
. level of liquid in the reactor tank is constant;
that is, the ow out is equal to the ow in. . Reaction rate coecient
. the reactor and the jacket are perfectly E

mixed. k k0 eRT273 32

The temperature controller is calibrated for a . Temperature transmitter


range of 80100 C. Table 2 shows the steady-state  
dTOt 1 Tt 80
and other operating information. TOt 33
The following equations constitute the process dt T 20
model.
. Sliding mode controller (SMCr)
. Mole balance on reactant A
Ut UC t UD t 34

dCA t Ft
CA t CA t kC2A t . Equal percentage control valve (air to close)
dt V
29 FC t FC max mt 35

Table 2
Design parameters and steady-state values

Variable Value Variable Value

CA 1.133 kgmol/m3 Vc 1.82 m3


Cai 2.88 kgmol/m3 Ft 0.45 m3/min
T 88 C Fc max 1.2 m3/s
Ti 66 C CPc 4184 J/kg- C
Tci 27 C 50
Set point 88 C T 0.33 min
HR 9.6e7 J/kgmol Ko 0.0744 m3/s-kgmol
CP 1.815e5 J/kgmol  C E 1.182e7 J/kgmol
U 3550.0 J/s m2  C Tc 50.5 C
c 1000 kg/m3 m 0.254 fraction CO
A 5.4 m2 V 7.08 m3
 19.2 kgmol/m3
Fig. 6. Scheme of continuous stirred tank reactor.
214 O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218
 
where t0  1 
mt m TOt TO l0 et
K t0 
CA t = concentration of the reactant in the KD St
reactor, kgmol/m3 18c
St 
CAi t = concentration of the reactant in the
feed, kgmol/m3
Tt = temperature in the reactor,  C with
Ti t = temperature of the feed,  C t
Tc t = jacket temperature,  C dTOt
St l1 et l0 etdt 18b
Tci t = coolant inlet temperature,  C dt 0
TOt = transmitter signal on a scale from 0 to 1
(fraction TO) With the values of K, , and t0 , the continuous
Ft = process feed rate, m3/s part of the SMCr can be tuned using the l
V = reactor volume, m3 expressions, Eqs. (19a) and (19b),
k = reaction rate coecient, m3/kgmol-s
HR = heat of reaction, assumed constant, l1 0:410 min1
J/kgmol
 = density of the reactor contents, l0 0:0421 min2
kgmol/m3
Cp = heat capacity of the reactants and
products, J/ kgmol-  C
U = overall heat-transfer coecient,
J/s-m2-  C
A = heat transfer area, m2
Vc = the jacket volume, m3
C = density of the coolant, kg/m3
Cpc = specic heat of the coolant, J/kg-  C
Fc t = coolant rate, m3/s
T = time constant of the temperature
sensor, s
Ut = SMCr output signal on a scale from
0 to 1 (fraction CO)
FCmax = maximum ow through the control
valve, m3/s
= valve rangeability parameter
k0 = Arrhenius frequency parameter,
m3/s-kgmol
E = activation energy of the reaction,
J/kgmol
R = ideal gas law constant, 8314.39
J/kgmol-K
mt = valve position on a scale from 0 to 1

Fig. 7 shows the open loop response of the


reactor; from this gure process parameters, are:
K 1:6 fraction TO/fraction CO;  13:0 min.;
t0 3:0 min. For this process, because the process
gain is positive, the SMCr is Fig. 7. Process reaction curve for the reactor.
O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218 215

Also , from Eqs. (19c) and (19d) never perfect. Martin [14] considers that modeling
error of 25% in its parameters is a ``reasonable
KD 0:96 fraction CO error''. Let us consider two cases. The rst case is
for 10% model error and the second one is for
 0:76 fraction TO=min 100% in model error. The second case could be
considered an ``unreasonable error,'' but our
Fig. 8 shows the system response when a +10% intent is to judge the controller. The error used is
change in inlet ow occurs. The gure shows that, the same in every parameter, that is, the same
because the temperature of the inlet ow is cooler 10% error in K,  and t0 .
than the temperature in the reactor, the reactor Fig. 9 shows the open loop responses for the
temperature rst decreases somewhat. However, actual process and for the model with a 10 and
after a short while the temperature in the reactor 100% error.
increases since more reactant is added to the reactor. Figs. 10 and 11 show the process response when
Fig. 8 shows the control performance when the the inlet ow changes by 10% and the modeling
modeling error between the real process and the error used is 10 and 100%, respectively. A com-
FOPDT model is small. However, the model is parison of Figs. 8 and 6, when no error in the

Fig. 8. System responses for 10% in inlet ow.


216 O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218

Fig. 9. Eect of modeling error.

Fig. 10. System responses for 10% change in inlet ow for 10% error in modeling.
O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218 217

Fig. 11. System responses for 10% change in inlet ow for 100% error in modeling.

model is present shows little dierence in the pro- The controller law, Eqs. (18a) and (18b) should
cess response. Fig. 9 shows that with 100% error be rather easy to implement in any computer sys-
in the model, the control performance degrades tem (DCS) [12].
somewhat. The most signicant dierence is that it
takes longer to return the process to the set point. References
However, even with such a large error in the
model, the control is still stable. [1] C.A. Smith, A.B. Corripio, Principles and Practice of
Automatic Process Control, John Wiley & Sons, New
York, 1997.
[2] V.I., Utkin, Variable structure systems with sliding modes,
6. Conclusions Transactions of IEEE on Automatic Control, AC 22
(1997) 212222.
This paper has shown the synthesis of a sliding [3] H. Sira-Ramirez, O. Llanes-Santiago, Dynamical dis-
mode controller based on an FOPDT model of the continuous feedback strategies in the regulation of non-
linear chemical processes, IEEE Transactions on Control
actual process. The controller obtained is of xed Systems Technology 2 (1) (1994) 1121.
structure. A set of equations obtains the rst esti- [4] J.J. Slotine, W. Li, Applied Nonlinear Control, Prentice-
mates for the tuning parameters. The examples Hall, New Jersey, 1991.
presented indicate that the SMCr performance is [5] M.C. Colantino, A.C. Desages, J.A. Romagnoli, A. Pala-
zoglu, Nonlinear control of a CSTR: disturbance rejec-
stable and quite satisfactory in spite of non-
tion using sliding mode control, Industrial & Engineering
linearities over a wide range of operating condi- Chemistry Research 34 (1995) 23832392.
tions. The relations given in Eq. (19) provided a [6] A.S.I. Zinober, Variable Structure and Liapunov Control,
good starting set of tunings. SpringerVerlag, London, 1994.
218 O. Camacho, C.A. Smith / ISA Transactions 39 (2000) 205218

[7] K.D. Young, V.I. Utkin, U. Ozgumer, A control engi- [11] O.E. Camacho, A new approach to design and tune slid-
neer's guide to sliding mode control, IEEE Transactions ing mode controllers for chemical processes, Ph.D. Dis-
on Control Systems Technology 7 (3) (1999) 328342. sertation, 1996, University of South Florida, Tampa,
[8] J.Y. Hung, W. Gao, J.C. Hung, Variable structure con- Florida.
trol: a survey, IEEE Transactions on Industrial Electro- [12] O. Camacho, C. Smith, E. Chacon, (1997). Toward an
nics 40 (1) (1993) 221. implementation of sliding mode control to chemical pro-
[9] G.E. Young, S. Rao, Robust sliding-mode of a nonlinear cesses, in: Proceedings of ISIE'97, Guimaraes-Portugal,
process with uncertainty and delay, Journal of Dynamical 1997, pp. 11011105.
Systems, Measurement, and Control 109 (1987) 202208. [13] D.M. Himmelblau, Applied Nonlinear Programming,
[10] O. Camacho, R. Rojas, W. Garcia, Variable structure McGraw-Hill, New York, 1972.
control applied to chemical processes with inverse [14] T.E. Marlin, Process Control, McGraw-Hill, New York,
response, ISA Transactions 38 (1999) 5572. 1995.

Oscar Camacho received the Electrical Engi- Carlos A. Smith, is Professor of Chemical Engi-
neering, and M.S. in Control Engineering degrees neering and Associate Dean for Academic Aairs
from Universidad de Los Andes (ULA)m Merida, at the University of South Florida. Before joining
Venezuela, in 1984 and 1992, respectively, and the USF he worked for Dow Chemical USA from
M.E. and Ph.D. in Chemical Engineering at Uni- 1967 through 1972. Professor Smith is a co-author
versity of South Florida (USF), Tampa, FL, in of two editions of Principles and Practice of Auto-
1994 and 1996, respectively. matic Process Control, 1986 and 1975, published
He has held teaching and research positions at by John Wiley & Sons. Professor Smith has been
ULA, CIED-PDVSA since 1985. He has been the in consultancy work, as well as lecturing on short
Chairman of the Electrical Engineering School at courses, for many companies in the US, Canada,
ULA since 1998. His current research interest Latin America, and Europe. Professor Smith loves
includes sliding mode control, hybrid systems and teaching, working with graduate students, and
passivity-based control. He is the author of more dealing with industry.
than 30 publications in journals and conference
proceedings.