Comparative Analysis of Different Controllers in Two–Area Hydrothermal Power System

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 04 Issue: 10 | Oct -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 493
Comparative Analysis of Different Controllers in Two–Area
Hydrothermal Power System
K. Kumar Swamy1, K. Chandra Sekhar2
1PG student, Dept. of EEE, Andhra University (A), Visakhapatnam, India
2Professor, Dept. of EEE, Andhra University (A), Visakhapatnam, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - In this paper we summarized the design and
implementation of fuzzy logic controller to solve automatic
power generation control problem in two-area hydrothermal
power system. The AGC performance is compared with
intelligent fuzzy logic control with conventional controllers
like PI, PID and PR under step load Disturbance. The
conventional controller Gains for PI and PID(kp, ki,kd) is
obtained by analyzing the transfer function using Ziegler
Nicholas Methods. The intelligent fuzzy controller simulation
is run to observe the performance of the system During 1%
step load disturbance. The simulation result show that the
fuzzy controller is better than the conventional PI, PID andPR
controllers in terms of Better Dynamic response and steady
error.
Key Words: Load Frequency control, Area Control
Error, controller Gain, PID, Ziegler NicholasMethod, PR,
Fuzzy logic control.
1. INTRODUCTION:
In a practical Interconnectedpowersystemconsists
of considerable number of generator, transmission line, Tie
Lines Loads… etc. For a continuous stable operation, A
Unpredictable change in load always cause power
generation-consumption mismatch which adversely affects
the quality of generatedpowerLikeFrequency, Voltage…. etc
A Automatic Generation Control scheme is to implemented,
it may Also Called as Load Frequency Control Because
Frequency is function of active power and voltage is a
function of reactive power. Therefore, two control loops are
used in power system. One is active power-frequency (P-f)
control loop. Second is reactive power-voltage (Q-V)control
loop. Attention of active power-frequency (P-f) control is
very important in comparison to reactive power-voltage(Q-
V) control because of mechanical inertia constant. The
control problem of the frequency and voltage can be
Decoupled [1]. Changing Generation has ConsiderableEffect
on The Frequency Compared to Voltage. The first ingenious
attempt is to control the frequency was via the flying wheel
governor of the synchronous machine. This governor's
action found to be insufficient and imposing a
supplementary control action turned out to be a necessity.
Supplementing the governor by a signal proportional to the
integral Controller of the frequency deviation from its
nominal value proved to be successful in achieving zero
steady state frequency deviation, but its dynamic
performance is unsatisfactory. Several attempts have been
done to enhance the performance ofTheSystembyDifferent
Classical and Numerical Methods Like Fuzzy Based Logic
Controller, An Artificial Neural Network [4], Variable
Structure Control, Meta-Heuristic Algorithms (MHAs),
Optimal Control Theory,LinerControllerFull StateFeedback
Control…etc. Classical approach based optimization for
controller gains is a trial and error method and extremely
time consuming [3].
When several parameters have to be optimized
simultaneously and provides suboptimal result. LFC is to
regulate by a signal called Area Control Error(ACE), which
accounts for errors in the interconnection frequency as well
as errors in the interchange power withfrequency,aswell as
errors in the interchange power with neighboring areas The
Main Aim of AGC is to Keep System Frequency and Tie Line
Power Exchange to Scheduled Value [2]. Theirvariations are
weighted together by a linear combination to a single
variable called the area control error (ACE).
Area control Error Acts as Input Signal for Designed
Controller. A control strategy is needed that not only
maintains constancy of frequency and desired tie-power
flow but also be able to achieve zero steady state error and
inadvertent interchange. AGC is the essential service in
maintainingthesystemintegritybymatchinggeneration and
demand in real time.
2. CONFIGURATION OF TWO AREA MODEL
Below Fig (1) Show the Configuration ofa TwoArea
Power System Connected Through a Tie-Line. Each Control
Area Consists of two or More Generator In each control area,
the generators are assumed to form a coherent group. Load
changes (ΔPd) at operating point affect both frequencies in
all areas and tie-line power flow between These he areas.
And the Equivalent is Given By
1 2
1 2
N
N
G G G G
H H H H
 
 

Where G is Generator Equivalent. And H is Inertia Constant.
Tie-Line
Area-1 Area-2
G1 G1
G2 G2
Gn Gn
Fig. (1)
3. SYSTEM MODELING
Thetwo-area interconnectedpowersystemtakenas
test system [4], In this study consists of thermal unit as
area-1and hydro unit as area-2. The control task is to
minimize the system frequency deviation 1f in area 1,
2f in area 2 and the deviation in the tie-line power flow
 Ptie between the two areas under the load disturbances
 Pd1 and  Pd2 in the two areas. This is achieved
conventionally with the help of integral control which acts
on iACE Given by (1), which is an input signal to the
controller where iACE the Area control error of the ith
area
,
1
n
i tie ij i i
j
ACE P B f

    (1)
 fi is Frequency error of ith area
 Ptie, i j are Tie-line power flow error between ith and jth
area
Bi is Frequency bias coefficient of ith area
1
1 g
ST
1
1 tST
1
1
1 ST
2
1
1
1
ST
ST


1
1 0.5
w
w
ST
ST


Turbine
TurbineHydro
/PI PR
Fuzzy
/PI PR
Fuzzy
1
11
p
p
K
ST
2
21
p
p
K
ST
02 T
S

1dP


  












 1
1 g
ST
1
1 tST
1
1
1 ST
2
1
1
1
ST
ST


1
1 0.5
w
w
ST
ST


Turbine
TurbineHydro
/P I P R
F u z z y
/PI PR
Fuzzy
1
11
p
p
K
ST
2
21
p
p
K
ST
02 T
S

2dP


 












/PI PR
Fuzzy
/PI PR
Fuzzy
Governer
IB
2B
1
1
R
2
1
R
2dP
1dP
Fig (2)
The Above figure show the interconnection of Basic Two
Power System, The Area Control Error of Each Unit is fed to
Different controller and Response is Observed.
4. DIFFERENT CONTROLLER
4.1 INTERGRAL CONTROLLER:
The conventional integral controllerisimplemented
and objective of any controller of load frequency is to
produce a controlling signal which keeps the frequency of
given system constant and power exchange betweencontrol
areas at predetermined values.Fig.1shows the typical
scheme of conventional control on ith control area. The area
control error(ACEi) is input to the PI controller with
proportional gain (kp) [5]
iK
iACE Control
Area
1dP
dif
tie
p
iB


i
u
Fig (3) shows the conventional controller
1B , 2B Are Frequency Bias Factors.
1dp is Change in Load In Area 1
12T is the Tie Line Constant depends Upon the System
Voltage of Two Control Area Connected Through the Tieline
and Its Reactance.
4.2 PID CONTROLLER
The PID controller design involves three separate
parameters, namely proportional, integral and derivative
gain values. The proportional action determinesthereaction
based on the current error, the integral action determines
the reaction based on the sum of recent errors and
derivative action determines the reaction based on the rate
at which the error has been changing, and the weighted sum
of these three actions is used to adjust the process via the
final control element. [5]. The transfer function of a PID
controller has the following Form

PK
iK
dK
PLANT




 U t
Figure (4)
( )
i
C S P d
K
G K K s
s
  
where kp, ki, kd are the proponational, integral and
derivative Gains and above equation Can Be written in
equivalent form of the PID Controllers
( )
1
(1 )S p d
i
G K T s
T s
  
Where / /i p i d d pT K K andT K K  , i dT andT are
Known as Integral and Derivative Time Constants
Respectively These controllers are used mostly in industrial
application because of simple implementation,morereliable
and easy realization. These controllers highly depend on
tuning parameters. This problem can be removed by most
popular Ziegler-Nichols method.
THE ZIEGLER-NICHOLS RULES (FREQUENCY RESPONSE
METHOD)
TABLE-1
Controller
pK iT iT
P 0.5ku
PI 0.4ku / 2uT
PID 0.6ku / 2uT / 8uT
Each Area Transfer Function is Given by
    
    1
1
1 1
( ) 2 2 1 1
g t
l
g t
s T s T ss
sp s s D T s T s K
R
  

     
TABLE-2
AREA
Ziegler Nichols tuning parameter
for simulation
Area-1 4.50, 1.73U UK T 
Area-2 0.45, 1.73U UK T 
Controller Gain
Area-1 PI
Controller
1.9226, 1.2p iK K  
Area-2 PI
Controller
0.1, 0.12p iK K  
Area-1 PID
Controller
2.70, 1.156, 0.216P i dK K K  
Area-1 PID
Controller
2.70, 1.156, 0.216P i dK K K  
Simulation is Done for Two – Area system for 0.01pu load
Change in Area-1 And Results Are Studied.
5. FUZZY LOGIC CONTROLLER
Fuzzy set theory and fuzzy logic establish the rules of a
nonlinear mapping. There has been extensive use of fuzzy
logic in control applications. Due to Non- Linearity of the
System Parameters, PI and PID controllers with fixed gain
parameters may not be provide better controlling of
frequency deviation in multi area power system. This
problem Can be solved by Using Fuzzy logic, Because the
Output of a controller is Self Tunned Depend upontheError.
Fuzzy logic controllers are especially used for control those
systems that are very typical to analyze by conventional
controller means they are not well defined by mathematical
formulation. One of its main advantages is that controller
parameters can be changed very quickly depending on the
system dynamics. [6]
The basic steps in modelling fuzzy based
controller after deciding the type of inference system
are:
 Fuzzyfication of crisp values
1.Extraction and normalization of crisp values for input
fuzzy vectors and output fuzzy vectors.
2. Selection of the membershipfunctions(MFs)-numberand
shape, for input fuzzy vectors and output fuzzy vectors.
3.Conversion of crisp values into Fuzzy inputs by calculating
membership grades[7].

 Rule base and Fuzzy Inference
1. Form a rule base using control observations.
2. Find out the rule bases that are stored
3. The rule base consists of easy to form simple if-then
conditional statementsthat decidethecontrol objectives and
control policy of the domain experts.
 Defuzzification
1. Calculate the crisp values for corresponding fuzzy output
vector, applying a suitable defuzzification process.
2. Results are procured after simulation.
For the proposed controller, the Mamdani fuzzy inference
engine was selected and realized by five triangular
membership functions for each of the three linguistic
variables (ACEi, d/dt(ACEi), Ki) with suitable choice of
intervals of the membership functions where ACEi and
d/dt(ACEi) act as the inputs of the controller and Ki is the
output of the controller. NB, NS, Z, PS, PB represent negative
big, negative small, zero, positive small, and positive big
respectively. [8]
TABLE 3
ACE
NB NS Z PS PB
 
d
ACE
dt
NB PB PB PB PS Z
NS PB PB PB Z Z
Z PS PS Z NS NS
PS Z Z NS NB NB
PB Z NS NB NB NB
The suitable choice of intervals of the Memberships
Functions was made as-0.1 to 0.1 for ACEi, -0.03 to 0.03 for
d/dt(ACEi)); and 0.001 to 1 for Ki
System performance in respect of deviations ⧍f1, ⧍f2 and
 Ptie for 0.01 p.u.MW step load change in area-1, using
Fuzzy Controller is studied [9].
6. PROPONATIONAL CONTROLLER
The Basic structure of proponational Controller is Shown in
Fig. (5)
()PRK s PLANT
( )US


Figure (5)
The PR current controller PRK is represented by:
  2 2PR p i
s
K s K K
s 
 

where, pK is the Proportional Gain term, iK is the Integral
Gain term and 0 is the resonant frequency. The ideal
resonant term on its own in the PR controller provides an
infinite gain at the ac frequency 0 and no phase shift and
gain at the other frequencies. [10]
7. RESULTS AND DISCUSSION
0 10 20 30 40 50 60 70 80 90 100
-2.5
-2
-1.5
-1
-0.5
0
0.5
x 10
-3
Time
ChangeinFrequency
Figure. (6). Change in Frequency in Area-1 with PI
Controller
0 10 20 30 40 50 60 70 80 90 100
-20
-15
-10
-5
0
5
x 10
-4
Time
ChangeinFrequency
Figure. (7) Change in Frequency ( 1)f in Area-2 With PI
Controller

0 10 20 30 40 50 60 70 80 90 100
-4
-3
-2
-1
0
1
2
x 10
-4
Time
ChangeinFrequency
Figure. (8) Change in Tie-Power with PI Controller
0 10 20 30 40 50 60 70 80 90 100
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
x 10
-4
Time
ChangeinFrequency
Figure. (9) Change in frequency ( 1)f in Area-1 With PR
Controller
0 10 20 30 40 50 60 70 80 90 100
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
1
x10
Time
ChangeinFrequency
Figure. (10) Change in frequency ( 1)f in Area-2 with PR
Controller
0 10 20 30 40 50 60 70 80 90 100
-1.5
-1
-0.5
0
0.5
1
1.5
2
x 10
Time
ChangeinTieLinepower
Figure. (11) Change in Tie Line Power with PR controller
0 20 40 60 80 100 120 140 160 180 200
-5
-4
-3
-2
-1
0
1
x 10
-3
Time
Changeinfrequency
Figure. (12) Change in frequency ( 1)f in Area-1 With
PID controller
0 50 100 150 200 250 300
-5
-4
-3
-2
-1
0
1
x 10
-3
Time
ChangeinFrequency
Figure. (13) Change in Frequency in Area -2 With PID
controller.

0 50 100 150 200 250
-10
-8
-6
-4
-2
0
2
4
x 10
-4
Time
ChangeinFrequency
Figure. (14) Change in Tie-line power with PID controller
0 10 20 30 40 50 60 70 80 90 100
-5
-4
-3
-2
-1
0
1
2
3
4
5
x 10
Time
ChangeinFrequency
Figure. (15) Change in Frequency in Area -1 With Fuzzy
Logic
0 10 20 30 40 50 60 70 80 90 100
-5
-4
-3
-2
-1
0
1
2
3
4
5
x 10
Time
ChangeinFrequency
Figure. (16) Change in Frequency in Area-2 With Fuzzy
logic
0 10 20 30 40 50 60 70 80 90 100
-5
-4
-3
-2
-1
0
1
2
x 10
ChangeinFrequency
Time
Figure. (17) Change in Tie-Line Power with Fuzzy Logic
0 10 20 30 40 50 60 70 80 90 100
-5
-4
-3
-2
-1
0
1
2
x 10
Time
ChangeinFrequency
Fuzzy
PI
PR
Figure. (18) Change in frequency with different controllers
8. CONCLUSIONS:
From the above tabulated and plotted simulation results
for the change in plant frequency and the tie line power, it is
clear that the intelligent Fuzzy based control minimizes the
settling time and maximum overshoot for change in system
frequency (f) and tie- line power that is for 1% change in
input power;1%change in frequency is observed
Proponational resonant controller Gives Steady state error
Minimum with Compromise of settling time. Thus the fuzzy
control methodology is faster and accurate as compared to
conventionally used PI, PID and PR controllers and Hence
steady state is achieved faster in case of Fuzzy logic
controllers for LFC of Two area system
9. APPENDIX
F=50 Hz, Ri=2.5 Hz/p.u. Megawatts, Tpi=20s, Tr=10s,
Hi=5 s, Kr=0.499, Pri==2000-Megawatt, Tt-i=0.299s,
Tgi=0.081s, Kpi=120 Hertz/p.u. Megawatts, Ki=4, Kd=5,
Tw=1 s, Di=8. 331*10 -3 p.u Megawatt/ Hz, Bi=0.4254 p.u
MegaWatt/Hz, ai=0.515, a= (2* pi*Ti ), del P di= 0.01,
Kp=0.05,Ki=-0.01, Kd=0.01.
10. REFERENCES
[1] O. I. Elgerd, Electric Energy Systems Theory: An
Introduction.
New York: McGraw-Hill, 1982.
[2] Nanda J, Kavi BL, “Automatic generation control of
interconnected power system,”IEEProceedings,Generation,
Transmission and Distribution, 1988; No. 125(5), pp.385–
390.
[3] Ibraheem, Prabhat Kumar and Kothari, D.P. (2005),
“Recent Phlosophies of AGC Strategies in Power Systems”,
IEEE Trans. On Pover System, Vol. 20, No. 1, pp. 346–357.

[4] D. K. Chaturvedi, P. S. Satsangi, and P. K. Kalra, “Load
frequency control: A generalized neural network approach,”
Elect. Power Energy Syst., vol. 21, no. 6, pp. 405–415, Aug.
1999
[5] Chang C.S., Fu W., “Area load-frequency control using
fuzzy gain scheduling of PI controllers,” Electric Power
System Research, Vol. 42, 1997, pp. 145-52.
[6] LEE, CC, “Fuzzy logic in control systems: Fuzzy logic
controller Part-I&II,” IEEE Trans. Syst., Man, Cybern.,vol. 20,
no. 2, pp. 404-435, March-April 1992.
[7] Chang C.S., Fu W., “Area load-frequency control using
fuzzy gain scheduling of PI controllers,” Electric Power
System Research, Vol. 42, 1997, pp. 145-52
[8] Sathans S., Swarup A., “Intelligent load frequency control
of two area interconnected power system and comparative
analysis,” IEEE Xplore, International conference on
communication systems and network technologies (CSNT
2011), pp. 360-365 E-ISBN 978-0-7695-4437-3.
[9]Sathans and A. Swarup, “Automatic GenerationControl of
Two-areaPower System with SMES: from Conventional to
Modern and Intelligent Control,” International Journal of
Engineering Science and Technology, vol. 3, no. 5, pp. 3693-
3707, 2011.
[10] L. H. Hassan, H. A. F. Mohamed, M. Moghavvemi, S. S.
Yang, “Automatic generation control of power system with
fuzzy gain scheduling integral and derivative controllers,”
International Journal of Power, Energy and Artificial
Intelligence, August 2008, No.1, Vol. 1.

Comparative Analysis of Different Controllers in Two–Area Hydrothermal Power System

More Related Content

What's hot (20)

Similar to Comparative Analysis of Different Controllers in Two–Area Hydrothermal Power System (20)

More from IRJET Journal (20)

Recently uploaded (20)

Comparative Analysis of Different Controllers in Two–Area Hydrothermal Power System