Load Frequency Control of Hydro-Thermal System with Fuzzy Logic Controller
Considering Boiler Dynamics
B. Anand, Member, IEEE,
Lecturer in Electrical Engg
Hindusthan College of Engg & Tech
Coimbatore, India – 641 032
E-mail: b_anand_eee@yahoo.co.in
Abstract- This paper describes the load frequency control
(LFC) of two area interconnected hydro-thermal system using
conventional proportional – integral (PI) controller and fuzzy
logic controller (FLC). The thermal system is incorporated with
governor dead band, generation rate constraint and boiler
dynamics. The hydro system is incorporated with generation
rate constraint. The conventional PI controller does not yield
adequate control performance with the consideration of non-
linearities and boiler dynamics. To overcome this drawback
FLC has been employed in the system. The aim of FLC is to
restore the frequency and tie line power in very smooth way to
its nominal value in the shortest possible time. Time domain
simulations are used to study the performance of the power
system. System performance is examined considering 1% step
load perturbation in either area of the system.
I. INTRODUCTION
Load Frequency Control is a very important issue in power
system operation and control for supplying sufficient and
reliable electric power with good quality. With an increasing
demand, the electric power system becomes more and more
complicated. For a successful operation of power system
under abnormal conditions, mismatches have to be corrected
via supplementary control. For satisfactory operation of a
power system the frequency should remain nearly constant.
The various areas or power pools are interconnected through
tie lines. These tie lines are utilized for contractual energy
exchange between power pools and also inter-area support in
case of abnormal conditions. The power system is subjected
to local variations of random magnitude and duration. As the
load varies at any area in the system considered, the
frequency related with this area affected and then the other
areas are also affected through tie lines. Frequency transients
must be eliminated as soon as possible. Mostly the boiler
system effects and its control, governor dead band effects and
generation rate constraint are neglected in the Load frequency
control studies for simplicity [1]. But for the realistic analysis
of system performance, these should be incorporated as they
have considerable effects on the amplitude and settling time
of oscillations.
The conventional proportional plus integral control
strategy, which is mostly used in power industry, take the
error signal as ACE. The conventional PI controller does not
provide adequate control performance with the consideration
of boiler dynamics and non-linearities such as governor dead
band and generation rate constraint. The difficulty in
A. Ebenezer Jeyakumar
Principal
Government College of Engg
Salem, India – 636 011
obtaining the optimum settling time of previously said
controller is mitigated by using FLC. Simulation results
confirm that the fuzzy logic controller greatly reduces the
overshoots. The settling time is also reduced considerably.
II. SYSTEM INVESTIGATION
The detailed block diagram of two area hydro-thermal
power system investigated in this study is shown in Fig.1 [1],
[17]. Area 1 comprising a reheat thermal system with
governor dead band, generation rate constraint non-linearities
and boiler dynamics also area 2 comprising a hydro system
with generation rate constraint. The nominal parameters of
the system are given in Appendix. Matlab version 7.3 has
been used to obtain dynamic response such as ΔF1, ΔF2, and
ΔPtie for 1% step load perturbation in either area of the
system.
A. Governor Dead Band
Governor Dead Band is defined as the total magnitude of a
sustained speed change within which there is no resulting
change in valve position. The Backlash non-linearity tends to
produce a continuous sinusoidal oscillation with a natural
period of about 2s [3], [8]. The speed governor dead band has
significant effect on the dynamic performance of load
frequency control system. Describing function approach is
used to incorporate the governor dead band non-linearity. The
hysteresis type of non-linearities are expressed as,
•
y = F ( x, x) rather than as y = F (x) (1)
To solve the non-linear problem, it is necessary to make
the basic assumption that the variable x, appearing in the
above equation is sufficiently close to a sinusoidal equation
that is,
x ≈ A Sin ω 0 t (2)
where, A is amplitude of oscillation
ω0 is frequency of oscillations
ω 0 = 2πf 0 = π
 As the variable function is complex and periodic function Drum type boiler is otherwise known as recirculation
of time, it can be developed in a Fourier series as follows, [3] boiler which relies on natural or forced circulation of drum
• N2 • liquid to absorb energy from the hot furnace walls, called
F ( x, x) = F 0 + N1 x + x + ....... (3) water walls for generating steam.
ω0 The boiler receives feed water which has been preheated in
As the backlash nonlinearity is symmetrical about the the economizer and provides saturated steam outflow.
0
origin, F is zero. For the analysis in this paper, backlash of Recirculation boiler make use of a drum to separate steam
approximately 0.05% is chosen [1]. From the above equation, flow from the recirculation water so that it can proceed to the
for simplification neglect higher order, the Fourier co- super heater as a heatable vapour; hence recirculation boiler
efficients are derived as N1=0.8 & N2=-0.2. By substituting are referred to as drum type boiler.
the values in Eq.(3) the transfer function for GDB is For boiler control strategies or modes of operations, here
expressed as follows, boiler-leading mode of control is considered. With this
operation the MW demand signal is applied to combustion
• 0 .2 •
F ( x , x ) = 0 .8 x − x (4) controls. Steam flow and MW output closely follow steam
π production in the boiler [11], [16].
B. Generation Rate Constraint
In practice, there exists a maximum limit on the rate of
change in the generating power. For thermal system a
generating rate limitation of 0.1 p.u per minute is considered,
[6], [18] i.e.
•
Δ P g ≤ 0.1 p.u. MW / min = 0.0017 p.u.MW / s (5)
For hydro system the rate of generation 4.5%/s is
considered [4].
C. Boiler Dynamics
Fig.2 shows the model to represent the boiler dynamics
[1], [4], [9]. Boiler is a device meant for producing steam
under pressure. The model is basically for a drum type boiler.
An oil/gas fired boiler system has been employed in this
study, since such boilers respond to load demand changes
more quickly than coal-fired units [10], [12].

Fig. 2 Boiler dynamics

III. CONVENTIONAL PI CONTROLLER

Literature survey shows that many utilities use PI

controller to achieve improved dynamic performance [14],
[17]. The task of Load frequency controller is to generate a
control signal (u) that maintains dynamic parameter at
predetermined values. The control signals can be written as,

u1 = − K p ⋅ ACE1 − K i ∫ ACE1dt (6)

u 2 = − K p ⋅ ACE2 − K i ∫ ACE2 dt (7)

Where, Kp and Ki are proportional and integral gains

respectively. For conventional PI controller, the gain Kp and
Fig. 1 Transfer function model of two area hydro-thermal system Ki has been optimized using Integral Square Error (ISE)
criterion. For ISE technique, the objective function used is,
 t
J = ∫ (ΔF12 + ΔF22 + ΔPtie2 ) dt
0
where,
∆F = change in frequency and
∆Ptie= change in tie line power
Using the ISE technique for PI controller, the optimum
value of Kp and Ki are found to be 0.002 and 0.02
respectively. Fig. 3(a)-(c) shows the responses for ΔF1, ΔF2,
and ΔPtie for 1% step load perturbation in hydro area. Nanda,
Mangla and Suri [5] have concluded that hydro area takes
more settling time than thermal area.
Fig. 3(a) Response with conventional PI controller (ΔF1)
Fig. 3(b) Response with conventional PI controller (ΔF2)
(8)
Fig. 3(c) Response with conventional PI controller (ΔPtie)
Examining the responses Fig. 3(a)-(c) it is seen that
frequency deviation in both area (ΔF1, ΔF2) and tie line
power deviation between area 1 and 2 (ΔPtie) are not properly
settled also overshoots are considerably high with
conventional PI controllers.
IV. FUZZY LOGIC CONTROLLER
The method of fuzzification has found increasing
applications in power systems. The applications of fuzzy sets
signify a major enhancement of power systems analysis by
avoiding heuristic assumptions in practical cases. This is
because fuzzy sets could be deployed properly to represent
power system uncertainties [15].
The design of FLC can be normally divided into three
areas namely allocation of area of inputs, determination of
rules and defuzzifying of output into a real value. In this
paper the proposed fuzzy controller takes the input as ACE
and ACĖ, which is given as, [13]
ACEi = ΔFi Bi + ΔPtie (9)
The parameter Bi may be optimized, but here, chosen as,
[2]
1 1
+ (10)
K p i Ri
The Block diagram of Fuzzy Logic Controller is shown in
Fig. 4. [7] Membership Function (MF) specifies the degree to
which a given input belongs to a set. Here, seven membership
function have been used to explore best settling time, namely
Negative Big (NB), Negative Medium (NM), Negative Small
(NS), Zero (ZO), Positive Small (PS), Positive Medium (PM)
and Positive Big (PB).
 Fuzzy rules are conditional statement that specifies the
relationship among fuzzy variables. These rules help us to
describe the control action in quantitative terms and have
been obtained by examining the output response to
corresponding inputs to the fuzzy controller. Rules are given
in Table I. The rules are interpreted as follows,

If ACE is NB and ACĖ is NS then output is PM.

Fig. 5(a) Frequency deviation in area 1

Fig. 4. Fuzzy Logic Controller

Defuzzification, to obtain crisp value of FLC output is

done by center of area method.

TABLE I
FUZZY RULES
ACE
NB NM NS ZO PS PM PB
NB PB PB PB PB PM PM PS
NM PB PM PM PM PS PS PS
NS PM PM PS PS PS PS ZO
ACĖ ZO NS NS NS ZO PS PS PS
PS ZO NS NS NS NS NM NM Fig. 5(b) Frequency deviation in area 2
PM NS NS NM NM NM NB NB
PB NS NM NB NB NB NB NB

V. SIMULATION RESULTS

Simulations were performed to a two area hydro-thermal

system considering governor dead band using describing
function approach, limits of the rate of generating power and
boiler effects for the thermal area and generation rate
constraint for hydro area. The system is simulated for a step
load disturbance of 1% on hydro area. Due to this the change
in dynamic response of the system has been observed, as
shown in Fig. 5(a)-(c). It is observed from the output
responses that the FLC minimized overshoots, improvement
in stability and settled the system properly. The Fig. 5(c) Tie line power deviation
conventional PI controller does not yield adequate control
with the consideration of boiler effects and non linearities. VI. CONCLUSION
Finally the simulation results of two-area system with FLC
have been compared with conventional PI control. Simulation In this paper FLC is designed for load frequency control of
result shows that the FLC yields much improved control two area interconnected hydro-thermal power system
performance when compared to conventional PI controller. considering non-linearities and boiler dynamics. The
conventional PI controller does not yield adequate control [4] A.Demiroren, and E.Yesil,”Automatic Generation Control with Fuzzy
Logic Controllers in the Power System Including SMES Units”,
performance with the consideration of non-linearities and Electrical Power and Energy Systems, 26 (2004) 291-305.
boiler effects. The simulation results conclude that FLC [5] Janardan Nanda, Ashish Mangla and Sanjay Suri, "Some New Findings
yields fast settling time which advocates the smooth on Automatic Generation Control of an Interconnected Hydrothermal
System with Conventional Controllers", IEEE Transactions On Energy
settlement of the quality power supply.
Conversion, Vol.21, No.1,March 2006, pp. 187-194.
[6] J.B.X.Devotta, M.G.Rabbani, and S.Elangovan, “Effects of SMES
Units on AGC Dynamics”, IEEE Catalogue No. 98EX137, pp. 61-66.
APPENDIX [7] J.Nanda and A.Mangla, “Automatic Generation Control of an
Interconnected Hydro-Thermal System using Conventional Integral and
Fuzzy Logic Controller,” 2004 IEEE International Conference on
(a) System data: [1], [4] Electric Utility Deregulation , Restructuring and Power Technologies,
Hongkong, April 2004, pp. 372-377.
Tt1 =0.3s [8] C.W.Taylor, K.Y.Lee and D.P. Dave,” Automatic Generation Control
Tg1=0.2s Analysis with Governor Deadband Effects”, IEEE Transactions on
Kr1=0.333s Power Apparatus Systems, Vol. PAS-98, No. Nov/Dec. 1979.
Tr1=10s [9] Kazuto Yukita, Yasuyuki Goto, Katsunori Mizuno, Toshihiro Miyafuji
and Katsuhiro Ichiyanagi, “ Study of Load Frequency Control using
Kp1=120 Hz/pu MW Fuzzy Theory by Combined Cycle Power Plant”, 2000 IEEE 0-7803-
Tp1=20s 6/00, pp. 422-427.
T12=0.0707 MW/rad [10] Xiangjie Liu and Tianyou Chai, “A Weighted Algorithm of Fuzzy
R1=R2 =2.4 Hz/ pu MW Logic Strategy on Water Level Control of Steam Generator”,
Proceedings of the 36th Conference on Decision & Control, San Diego,
Kp2=80 Hz/pu MW USA,Dec. 1997, pp. 3357-3362.
Tp2=13s [11] F.Laubli and F.M. Fenton,Jr., “ The flexibility of the Supercritical
T1=48.7s (hydro area) Boiler as a Partner in Power System Design and Operation. Part I :
T2=0.513s (hydro area) Theoretical Relationships”, IEEE Winter Power Engineering Society
Meeting, January 31-February5, 1971,N.Y. pp. 1719-1724
T3=10s (hydro area) [12] C.S.Chen, Y.D.Lee and C.K.Chen,”Transient Response of Incinerator
Tw=1s (hydro area) Plant by Considering Boiler Model with Dynamic Steam Variation”,
f=60 Hz IEEE Transactions on Power Systems, Vol.23, No.1, February 2008.
B1=B2=β=0.425 [13] J.Nanda and J.S.Sakkaram,” Automatic Generation Control with Fuzzy
Logic Controller Considering Generation Rate Constraint”,
Proceedings of 6th International Conference on APSCOM 2003,
Hongkong, November2003.
(b) Boiler (oil fired) data: [1] [14] Ertugrul Cam and Ilhan Kocaarslan “A Fuzzy Gain Scheduling PI
Controller Application for an Interconnected Electrical Power System”,
Electrical Power System Research 73 (2005)267-274.
K1= 0.85 [15] Timothy J.Ross, “Fuzzy Logic with Engineering Applications”,
K2= 0.095 McGraw Hill International Edition
K3= 0.92 [16] Kundur.P,” Power System Stability and Control”, Mc Graw Hill, New
Cb= 200 York 1994.
[17] C.S.Chang and Weihui Fu, “Area Load Frequency Control using Fuzzy
T d= 0 Gain Scheduling of PI Controllers”, Electrical Power System Research
Tf= 10 42 (1997) 145-152.
Kib= 0.03 [18] C.T.Pan and C.M.Liaw, “An adaptive controller for power system load
Tib= 26 frequency control”, IEEE Transactions on Power Systems, Vol. 4,
No.1, February 1989.
Trb= 69

REFERENCES

[1] S.C.Tripathy, R.Balasubramanian and P.S.Chandramohanan Nair,

"Effect of Superconducting Magnetic Energy Storage on Automatic
Generation Control Considering Governor Deadband and Boiler
Dynamics",IEEE Transactions on Power Systems, Vol.7, No.3, August
1992, pp. 1266–1272.
[2] Aysen Demiroren, Neslihan S.Sengor, and H.Lale Zeynelgil,
"Automatic Generation Control By Using ANN Technique", Electric
Power Components and Systems,29:883-896, 2001, pp. 833-895.
[3] S.C.Tripathy, T.S. Bhatti, C.S. Jha, O.P.Malik and G.S.Hope, "Sampled
Data Automatic Generation Control Analysis with Reheat Steam
Turbines and Governor Dead-Band Effects", IEEE Transactions on
Power Apparatus and Systems, Vol.PAS-103, No.5, May 1984, pp.
1045-1050