INTERNATIONAL

JOURNAL OF
International Journal of Engineering, Science and Technology ENGINEERING,
MultiCraft Vol. 3, No. 4, 2011, pp. 14-26 SCIENCE AND
TECHNOLOGY
www.ijest-ng.com
© 2011 MultiCraft Limited. All rights reserved

Decentralized controller gain scheduling using PSO for power

system restoration assessment in a two-area interconnected
power system

R. Jayanthi1 *, I. A. Chidambaram2, C. Banusri3

1*
Department of Electrical Engineering, Annamalai University, Annamalainagar -608002, INDIA
2
Department of Electrical Engineering, Annamalai University, Annamalainagar -608002, INDIA
3
Department of Electrical Engineering, Jaya College of Engineering, Avadi, Chennai. INDIA
*
Corresponding Author: e-mail: [email protected], Tel +91-4144-237092

Abstract

Load Frequency Control (LFC) is one of the most important issues in electrical power system design/operation and is
becoming much more significant recently with increasing size, changing structure and complexity in restoration of
interconnected power systems. In practice, LFC systems are used with simple Proportional-Integral (PI) or Integral (I)
controllers. However, since the PI or I control parameters are usually tuned based on classical or trial-and-error approaches, they
are incapable of obtaining good dynamic performance if the power system is more vulnerable due to various load changing
scenarios in multi-area power system. For this reason, in this study the P and I control parameters are tuned based on Particle
Swarm Optimization (PSO) algorithm for a better Load-Frequency Control in a Two-Area Two-Unit Thermal Reheat Power
System (TATURIPS) with step load perturbation. To exemplify the optimum parameter search PSO is used as it is an
optimization method, therefore, even in the uncertainty area of controller parameters, finds the best parameters for controller and
obtained controller is an optimal controller. This makes a trade-off between exploration and exploitation of search space to find
global optimum in less number of generations. A TATURIPS is taken for the study to illustrate the proposed method. To show
effectiveness of proposed method, the performance of optimized PI controller is obtained with several time-domain simulations
for various load changes scenarios and is presented. Simulation results emphasis on the better settling time based stability
performance of optimized PI controller in the TATURIPS with GT unit when compared with that of the SMES and the
conventional system two-unit two-area interconnected power systems.

Keywords: Load Frequency Control, Particle Swarm Optimization, TATURIPS, Gas Turbine, Super Conducting Magnetic
Energy Storage Device.

1. Introduction

Large-scale power systems are normally composed of interconnected subsystems or control areas. The connection between the
control areas is done using tie-lines. Each area has its own generator or group of generators and it is responsible for its own load
and scheduled interchanges with neighboring areas. Because loading of a given power system is never constant and to ensure the
quality of power supply, a load frequency controller is needed to maintain the system frequency at the desired nominal value. It is
known that changes in real power affect mainly the system frequency and the input mechanical power to generators is used to
control the frequency of the output electrical power. In a conventional power system, each control area contains different kinds of
uncertainties and various disturbances due to increased complexity, system modeling errors and changing power system structure.
A well designed power system should not only cope up with changes in load but also with system disturbances and it should
provide acceptable high level of power quality while maintaining both voltage and frequency within tolerable limits (Wang Y, et
al, 1998). During the last three decades, various control strategies for LFC have been proposed. This extensive research is due to
the fact that LFC constitutes an important function on power system operation where the main objective is to regulate the output
power of each generator at prescribed levels while keeping the frequency fluctuations within pre-defined limits. Robust adaptive
15 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

control schemes have also been developed (Yamashita. et al, 1991, Pan et al., 1989, Aldeen et al., 1991) to deal with changes in
system parametric under LFC strategies. Viewing a multi-area power system under LFC as a decentralized control design for a
multi-input multi output system, it has been shown that a group of local controllers with tuning parameters can guarantee the
overall system stability and performance. The result reported in Lim et al. (1996) demonstrates clearly the importance of
robustness and stability issues in LFC design. The applications of artificial neural networks, genetic algorithms, fuzzy logic and
optimal control to LFC have been reported in (Shayegi et al, 2007, Shayeghi. et al., 2007, Chao ou, et al, 2006). The objective of
this study is to investigate the load frequency control, of a multi-area power system taking into consideration of various level of
step sized loading to assess the requirements for power system restoration. An optimal control scheme based particle swarm
optimization (PSO) Algorithm method is used for tuning the parameters of PI controller (S.P. Ghoshal, 2004). The proposed
controller is simulated for a two-area two-unit thermal reheat power system and the output responses of the system with respect to
the settling time has been studied. To show the effectiveness of proposed method and also to compare the performance of these
two controllers, several changes in demand of first area, demand of second area and demand of two areas simultaneously are
applied. Simulation results indicate that controllers designed using PSO approach guarantees good performance under various load
conditions. The expert system, which is used to bring up the vulnerable power system to a normal system which allows the
estimation and observation of the real restoration time, the degree of stability, the observation of the system voltage profile, power
to be transmitted can be done using the following tools (Adibi et al., 1994).

1.1Generation Management: This tool is responsible for making the generators to commit for the optimal generation. Firstly, it
started by finding the smallest black start generator in the solution and then it connects the generators in accordance with the
generator sequence provided by the PSO solution.
1.2Restoration Path Management: In every step of connecting a generator or load, an optimized path algorithm is used to find the
shortest path. Moreover the Path Management is used to check the loading limits of every line proposed for connection.
1.3 Time Management: Since one of the main goals of using the expert system is to estimate the real restoration time, great
attention has been given to the time required for every element in every stage of restoration.
1.4 Load management: During restoration, loads are restored based on the load priorities and system security considerations.
The priorities of loads are calculated in accordance to the load importance. If two loads are in the same degree of priority, the
nearest one is picked. Moreover, if two or more loads are in the same degree of priority and in the same distance, the load with
the highest level of connectivity is picked.
1.5 Role of expert system in this proposed work: The main objectives of expert system in knowledge based restoration are

• By providing initial source of power immediately to the interconnected power system with SMES unit.
• By optimizing the gain values of the PI controller using PSO technique for the two-area interconnected power system
with SMES unit for system restoration.
• The primary function of the expert system is to restore the interconnected thermal reheat power system even for small
disturbances and to avoid excess under frequency deviations.

2. Modeling of a two-area interconnected thermal reheat power system

Due to the inherent characteristics of changing loads, the operating point of power system may change very much during a daily
cycle. The generation changes must be made to match the load perturbation at the nominal conditions, if the normal state is to be
maintained. The mismatch in the real power balance affects primarily the system frequency but leaves the bus voltage magnitude
essentially unaffected. In a power system, it is desirable to achieve better frequency constancy than obtained by the speed
governing system alone. This requires that each area should take care of its own load changes, such that schedule tie power can be
maintained. A two-area interconnected system dynamic model in state variable form can be conveniently obtained from the
transfer function model. Figure 1 represents the transfer function model of a two-area reheat thermal system.
The state variable equation of the minimum realization model of the ‘N’ area interconnected power system is expressed as
(Chidambaram et al., 2005).
•
X = Ax+ Bu +Γd (1)
Y = Cx (2)
Where, the system state vector x consists of the following variables as:

[x] = [∫ ACE1dt , ∫ ACE2 dt , ΔF1 , ΔPg1 , ΔX e1 , ΔPtie , ΔF2 , ΔPg 2 , ΔX e 2 ]

T

u = [u1,…., uN]T = [ΔPc1,…,ΔPcN]T N – Control input vector

16 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

d = [d1,….dN]T = [ΔPD1,….,ΔPDN]T N – Disturbance input vector

y = [y1,….,yN]T 2N – measurable output vector

A is system matrix, B is the input distribution matrix and Γ disturbance distribution matrix, x is the state vector, u is the control
vector and d is the disturbance vector of load changes of appropriate dimensions.

β1 Controller 1 ΔPD1(S)
R1

+ _ _
1 1 1 + K r1Tr1s K p1 ΔF1(s)
K 11 _
+ U1 1 + sTg1 1 + sTt1 1 + Tr1s + 1 + sT p1
s
ACE1 Governor Turbine
ΔXE1(s) ΔPr1(s) Reheat ΔPg1(s) ΔPtie1 (s) Power system
+
2π T12
S _
a12 Controller

ΔXE2(s) ΔPr2(s) a12

ACE2 ΔPg2(s) ΔPtie 2 (s)
+
U2 _
K12 1 1 1 + K r 2Tr 2 s + K p2
+
s _ _ 1 + sTg 2 1 + sTt 2 1 + Tr2 s _ 1 + sT p 2 ΔF2(s)
β2
1
Governor Turbine Reheat ΔPD2(S) Power system

R2

Figure 1. Transfer function model of two-area reheat thermal system

The typical values of system parameters for nominal operation condition are given in appendix. This study focuses on optimal
tuning of controllers for LFC and tie-power control with settling time based optimization using PSO algorithm to ensure a better
power system restoration assessment. The aim of the optimization is to search for the optimum controller parameter setting that
maximizes the minimum damping ratio of the system. On the other hand in this study the goals are control of frequency and inter
area tie-power with good oscillation damping and also obtaining a good performance under all operating conditions and various
loads and finally designing a low-order controller for easy implementation. The gain values and their respective settling time for
both 1% and 4% in area1 and area 2 are given in tables 2,3,4 & 5 respectively.

3. Modeling of a Gas turbine

Amid growing concerns about Green house emissions, Gas turbines have been treated as a viable option, due to their higher
efficiency and the lower green house gas emissions compared to other energy sources and fast starting capability which enables
them to be often used as peak units that respond to peak demands(Soon klat yee et al, 2008). Also, they can be profitably used in
power system restoration for supplying power to the restoration areas as they have the advantages like, Quick start-up/shut-down,
Low weight and size, Cost of installation is less, Low capital cost, Black-start capability, High efficiency, Requires low cranking
power, Pollutant emission control etc.,
The continuous power plant output of a Gas turbine at the maximum depends upon frequency and temperature. It gives
approximately two-thirds of the total power output of a typical combined cycle plant (S. Barsali, et al 2008, M. Nagpal, et al,
2001). When the load is suddenly increased the speed drops quickly, but the regulator reacts and increases the fuel flow to a
maximum of 100% thereby improving the efficiency of the system. A model as shown in fig 2 has been incorporated in the system
under study.
17 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

SPEED
Dturb

1
R
vmax
-
+ Low + -
1 1
Load Σ value Σ PMECH
Reference
gate 1 + T1 s 1 + T2 s
VAR(L)
vmin

+ - 1
Σ KT Σ
+ 1 + T3 S
+

Load limit
Figure 2. Gas Turbine Model

4. Super Conducting Magnetic Energy Storage (SMES) device

The normal operation of a power system is continuously disturbed due to sudden small load perturbations. The problem lies in
the fact that the inertia of the rotating parts is the only energy storage capacity in a power system. Thus, when the load-end of
the transmission line experiences small load changes, the generators need continuous control to suppress undesirable oscillations
in the control to suppress undesirable oscillations in the system. The superconducting magnetic energy system is a fast acting
device which can swallow these oscillations and help in reducing the frequency and tie-line Power deviations for better
performance of system disturbances. The SMES is designed to store electric energy in the low loss superconducting coil. Power
can be absorbed or released from the coil according to the system requirement. A SMES which is capable of controlling active
and reactive power simultaneously has been expected as one if the most effective stabilizers for power oscillations (A.
Demiroren 2002). Besides oscillation control, a SMES allows a load leveling, a power quality improvement and frequency
stabilization. A typical SMES system includes three parts namely superconducting coil, power conditioning system and cooled
refrigerator. From the practical point of view, a SMES unit with small storage capacity can be applied not only as a fast
compensation device for power consumptions of large loads, but also as a robust stabilizer for frequency oscillations.

4.1. SMES UNIT:

The schematic diagram in Fig. 3 shows the configuration of a thyristor controlled SMES unit (S.C. Tripathy, et al, 1997). The
SMES unit contains DC superconducting Coil and converter which is connected by Y–D/Y–Y transformer. The inductor is
initially charged to its rated current Id0 by applying a small positive voltage. Once the current reaches the rated value, it is
maintained constant by reducing the voltage across the inductor to zero since the coil is superconducting (H. Shayegi, et al,
2010). Neglecting the transformer and the converter losses, the DC voltage is given by

Ed= 2Vd0 cos α- 2IdRc (3)

Where Ed is DC voltage applied to the inductor, firing angle (α), Id is current flowing through the inductor. Rc is equivalent
commutating resistance and Vd0 is maximum circuit bridge voltage. Charge and discharge of SMES unit are controlled through
change of commutation angle α (A. Demiroren, et al, 2004).
18 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

Id

AC
System Ed Superconducting
L
Bus Inductor
Y – Y/Δ
Transformer

12 pulse bridge
Converter
Figure 3. The schematic diagram of SMES unit

In AGC operation, the dc voltage Ed across the superconducting inductor is continuously controlled depending on the sensed
area control error (ACE) signal. Moreover, the inductor current deviation is used as a negative feedback signal in the SMES
control loop. So, the current variable of SMES unit is intended to be settling to its steady state value. If the load is used as a
negative feedback signal in the SMES control demand changes suddenly, the feedback provides the prompt restoration of
current. The inductor current must be restored to its nominal value quickly after a system disturbance, so that it can respond to
the next load disturbance immediately. As a result, the energy stored at any instant is given by

WL=LId2/2 MJ (4)

Where L=inductance of SMES, in Henry

Equations of inductor voltage deviation and current deviation for each area in Laplace domain are as follows

⎛ K SMES ⎞
⎟ [β 1 Δ F1 ( s ) + Δ Ptie 1 ( s ) ] −
K id
∆Edi(s) = ⎜ Δ I di ( s ) (5)
⎝ 1 + sT dci ⎠ 1 + sT dci

∆Idi(s) = (1/sLi) *∆ Edi(s) (6)

Where
∆Edi(s) = converter voltage deviation applied to inductor in SMES unit
KSMES = Gain of the control loop SMES
Tdci = converter time constant in SMES unit
Kid = gain for feedback ∆Id in SMES unit.
∆Idi(s) = inductor current deviation in SMES unit
The deviation in the inductor real power of SMES unit is expressed in time domain as follows

∆PSMESi=∆EdiIdoi+∆Idi∆Edi (7)

Figure 4 shows the block diagram of the SMES unit. To achieve quick restoration of the current, the inductor current deviation can
be sensed and used as a negative feed back signal in the SMES control loop.
19 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

Ied

KM

1 1
ACE Ks Σ Σ
1 + sTds Ls ΔIs
IsΔIs

Σ
Figure 4. Block diagram of SMES unit ΔPs

In a two-area interconnected thermal power system under study (as shown in Fig 5) with the sudden small disturbances which
continuously disturb the normal operation of power system.

U smes P smes

SMES

0.425

BIAS FACTOR

0.416

SPEED REGULATION
1
Scope 1
In 1 Product 1 1 5s+1
120
0.08 s+1 0.3s+1 10 s+1
20 s+1
GOVERNOR TURBINE REHEAT
POWER SYSTEM
1 1 5s+1
PID
0.08 s+1 0.3s+1 10 s+1
PID CONTROLLER GOVERNOR 1 TURBINE 1 REHEAT 1
SPEED REGULATION 1 LOAD

0.416
0.545
s
Synchronizing coefficient
0.425 -1
-1 AREA1 to 2
AREA 1to 2
BIAS FACTOR 1

0.416

PID CONTROLLER 1 120

SPEED REGULATION 2
20 s+1
PID 1 1 5s+1
POWER SYSTEM 1
0.08 s+1 0.3s+1 10 s+1
GOVERNOR 2 TURBINE 2 REHEAT2

1 1 5s+1
LOAD1
0.08 s+1 0.3s+1 10 s+1
GOVERNOR 3 TURBINE 3 REHEAT 3

0.416

SPEED REGULATION 3
In1Out 1

GAS TURBINE

Figure 5. Simulink model of TATURIPS with SMES and Gas Turbine units
20 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

As a result the requirement of frequency controls of areas beyond the governor capabilities SMES is located in area1 absorbs
and supply required power to compensate the load fluctuations (Issarachi Nagamroo, 2005, A.Demiroren, et al, 2003, A.
Demiroren et al 2004).
Tie-line power flow monitoring is also required in order to avoid the blackout of the power system. The Input of the integral
controller of each area is
ACEi= βi∆fi+∆Ptie i (8)
Where,
βi= frequency bias in area i
∆fi = frequency deviation in area i
∆Ptie i= Net tie power flow deviation in area i

The application of energy storages to electrical power system can be grouped into two categories.
1. like conventional pumped hydro plant storage meant for load leveling application.
2. To improve the dynamic performance of power system.
SMES have the following advantages like: The time delay during charge and discharging is quite short, Capable of controlling
the both active and reactive power simultaneously, Loss of power is less, High reliability, High efficiency.

5. Controller design using particle swarm optimization technique for the power system restoration problem

• This is a population based search technique

• Each individual potential solution in PSO is called particle.
• Each particle in a swarm fly around in a multidimensional search space based on its own experience and experience by
neighboring particles
• Let in search space ‘S’ in n-dimension with the swarm consists of ‘N’ particles
Let, at instant‘t’, the particle ‘i’ has its position defined by
{
X ti = x1i , x 2i , . . . x ni }
i
{
Velocity V t = v1 , v 2 ,
i i
}
. . .v ni in variable space ’S’
• Velocity and position of each particle in the next generation (time step) can be calculated as
( ) (
V t i+1 = ω V t i + C 1 . rand ( ) . Pt i − X ti + C 2 Rand ( ) Pt g − X ti )
X i
t +1 = X +V t
i i
t +1
where N – number of particle in swarm
ω - inertia weight
C 1 , C 2 - acceleration constant
rand ( ) Rand ( ) - Uniform random value in the range [0, 1]
Pt i - best-position that particle ’i’ could find so far
Pt g - global best at generation‘t’
• Performance of PSO depends on selection of inertia weight ( ω ), Max velocity V max and acceleration constant
( C1 , C 2 )
™ Effect of
ƒ Inertia weight ( ω )
• Suitable weight factor helps in quick convergence
• Large weight factor facilitates global exploration (i.e. searching of new area)
• While small weight factor facilitates local exploration (so wise to choose large weight factor for initial
iterations and gradually smaller weight factor for successive iterations)
• Generally, ω 0.9 at beginning and 0.4 at end (Y. Shi et al, 1998)
ƒ Max velocity V max
• With no restriction on the max velocity of the particle, velocity may become infinitely large.
• If V max is very low particle may not explore sufficiently
21 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

• If V max is very high it may oscillate about optimal solution. Therefore, velocity clamping effect has to be
introduced to avoid ‘swarm explosion’ (J. Kennady et al, 2001)
• Generally, max velocity is set as 10-20% of dynamic range of each variable.
• Velocity can be controlled within a band

V ini − V fin
V max = V ini − iter
iter max
ƒ Acceleration constant ( C 1 , C 2 )
• C1 is called Cognitive Parameter which pulls each particle towards local best position.
• C1 , C 2 is called Social Parameter which pulls the particle towards global best position.
• Generally, C1 , C 2 are chosen between 0 to 4
The design steps of PSO based PI controller is as follows.
1. The algorithm parameters like number of generation, population, inertia weight and constants are initialized.
2. The values of the parameters KP and Ki initialized randomly.
3. The fitness function of each particle in each generation is calculated.
4. The local best of each article and the global best of the particles are calculated.
5. The position, velocity, local best and global best in each generation is updated
6. Repeat the steps 3 to 5 until the maximum iteration reached or the best solution is found.

5.1 Simulink model of PSO Based PI Controller

Figure 6. shows the simulink model of the plant with PSO algorithm based PI controller.The PSO algorithm is used to search an
optimal parameter set containing Kp and Ki . The parameters used for tuning the PSO algorithm and simulink models are tabulated
in table below and Fig. 7 represents the corresponding flow chart:

SCOPE
PSO BASED PI CONTROLLER PLANT
STEP Vref
+ PI LFC
-

1
OUT 1
1 *2
V2
s
SQUARE INTEGRATOR WORKSPACE
FUNCTION
Figure 6. Simulink model of plant with PSO Algorithm based PI Controller
Table 1. Parameter values tuned for PSO algorithm
Parameters Value
Population size 5
Number of generations 10
Inertia weight ( ω ) 0.8
cognitive coefficient (C1) 2.05
social coefficient (C2) 2.05
22 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

Start

Select the parameters of PSO: population size,

Maximum no. of iter c1, c2 and weights

Initialization of components (PI gains), each of Particles

With random positions &velocities

Iteration=1 and weight updating

Evaluate the value fitness of each Particle in current position

Update pbest and gbest of each particle

Optimal values of PI
controller

Is iter: max Iter? Yes

End

No

Update particles positions and velocity using equations

velocity=w*velocity+c1 (R1*local best position-current position) +c2
(R2*global best position-current position) current position=current position
+ l it

Iter=iter+1

Figure 7: Flowchart for PSO

Table 2. Proportional plus Integral controller gains for 0.01p.u. MW step-load change in Area-2
Power system Gain values
KP KI
Conventional in Area1 & 2 0.95 0.30
With SMES in Area-1 0.52 0.26
With GT in Area-2 0.65 0.24
23 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

CASE 1: Comparison of frequency deviations, control input requirements and tie-line power deviations in a two-area
interconnected thermal reheat power system for 0.01p.u. MW load change in area-2.

Δf1 in Hz

Time in seconds (a)

Δf2 in Hz
ΔPc1 in p.u MW

Time in seconds (b)

ΔPc2 in p.u MW

Time in seconds (c)

ΔPtie in p.u MW

Time in seconds (d)

Time in seconds (e)

(a) Frequency deviation without and with SMES
(b) Frequency deviation without and with Gas Turbine
(c) Control input requirement without and with SMES
(d) Control input requirement without and with Gas Turbine
(e) Tie-line power deviations without and with Gas Turbine

Table 3: Settling Time (in seconds) of the output response without and with GT For 0.01p.u. MW step load change in area-2
Power System Change in frequency H.Z ∆Ptie1

Without GT More than 45 sec. (∆F2) More than 50 sec

With GT in area-2 13 sec. (∆F2) 12 sec
Without SMES More than 50 sec. (∆F1) More than 50 sec
With SMES in area-1 25 sec. (∆F1) 13 sec.
24 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

Table 4: Proportional plus Integral controller gains for 0.04 p.u. MW step-load change in Area-2
Power system Gain values
KP KI
Conventional in Area 1 & 2 0.98 0.22
With SMES in Area-1 0.64 0.15
With GT in Area-2 0.7 0.26

CASE 2: Comparison of frequency deviations, control input requirements and tie-line power deviations in a two-area
interconnected thermal reheat power system for 0.04p.u. MW load change in area-2.
Δf1 in Hz

Time in seconds (a)

Δf2 in Hz

…… Without GT
____With GT

Time in seconds (b)

ΔPc1 in p.u MW

Time in seconds (c)

ΔPc2 in p.u MW

…… Without GT
____With GT

Time in seconds (d)

ΔPtie in p.u MW

…… Without GT
____With GT

Time in seconds (e)

(a) Frequency deviation without and with SMES
(b) Frequency deviation without and with Gas Turbine
(c) Control input requirement without and with SMES
(d) Control input requirement without and with Gas Turbine
(e) Tie-line power deviations without and with Gas Turbine

Table 5: Settling Time (in seconds) of the output response without and with GT For 0.04p.u. MW step load change in area-2

Power System Change in frequency H.Z ∆Ptie1

Without GT More than 45 sec. (∆F2) More than 50 sec

With GT 12 sec(∆F2) 13 sec
Without SMES More than 50 sec. (∆F1) More than 50 sec
With SMES 35 sec. (∆F1) 18 sec.
25 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

6. Conclusion

In this work a novel heuristic search technique (i.e.) PSO based controllers are designed and implemented in a two-area two-unit
thermal reheat power system with GT in area 2 and SMES in area 1 by optimizing the gain values of the PI controller to restore the
frequency and tie-line power deviations due to various step-load changes in area 2. The output response of the system reveals that
reduced peak overshoots, minimum settling time are obtained thereby maintaining the stability of the system under study to greater
extent when comprised with GT unit. Using this global optimization controller the improvement of the dynamic performance of
TATURIPS is found to be very clear in attaining the restoration of the system under consideration. The proposed controller yields
good transient response with a minimum settling time. Further, this work is being extended with the consideration of the system
non-linearities.

APPENDIX
Data for the two-area interconnected thermal power system with reheat turbines (Chidambaram I.A et al., 2005)
Pr1=Pr2=2000MW
Kp1=Kp2=120Hz/p.u
Tp1=Tp2=20sec.
Tt1=Tt2=0.3 sec.
Tg1=Tg2=0.08sec.
Kr1=Kr2=0.5
Tr1=Tr2=10 sec.
R1=R2=2.4Hz/p.u MW.
a12=-1
T12=0.545 p.u MW/Hz
β 1 = β2 =0.425 p.u. MW/Hz
Data for the SMES unit (Demiroren.A 2002)
L=2.65H
Tdc=0.03 sec
Ido=4.5KA
Kid=0.2 KV/KA
KSMES=100KV/unit MW
Data for the Gas turbine model (Soon klat yee et al., 2008)
T1=10 sec
T2=0.1sec
T3= 3sec
Kt=1
Kr=0.04
Dturb=0.03
Maximum and minimum valve position = 1and -0.1

Acknowledgement
The authors wish to thank the authorities of Annamalai University, Annamalai Nagar, Tamilnadu, India for the facilities provided
to prepare this paper.

References

Adibi M.M., R.J Kafka,D.P Milanic,Aug 1994 Expert system requirements for power system Restoration, IEEE Transactions on
Power Systems, Vol.9, pp.1592-1598.
Aldeen, M. and J.F. Marsh, 1991. Decentralized PI design method for interconnected power systems. IEE Proc.-C, Vol. 138,263-
274, No. 4.
Barsali S., D. Poli, A.Pratico, R.Salvati, M.Sforna, Aug 2008, Restoration islands supplied by gas turbine, Electrical Power System
Research, Vol. 78, pp. 2004-2010.
ChaoOu, Weixing Lin, June 2006, Comparison between PSO and GA for parameters optimization of PID Controller, Proceedings
of the IEEE International Conference on Mechatronics and Automation, pp. 2471-2475.
Chidambaram I.A. and S.Velusami, 2005, Design of decentralized biased controllers for load-frequency control of interconnected
power systems, Electric Power Components and Systems, Vol. 33, No.12, pp.1313-1331.
26 Jayanthi et al. / International Journal of Engineering, Science and Technology, Vol. 3, No. 4, 2011, pp. 14-26

Demiroren A., 2002 Application of a self-tuning to power system with SMES, European Transactions on Electrical Power
(ETEP), Vol. 12, No. 2, pp. 101-109.
Demiroren A., 2003 AGC for power system with SMES by using NN controller, Electric Power Components and Systems, Vol.
31, No. 1, pp. 1-25.
Demiroren A., HL Zeynelail Sengor, 2004, AGC using ANN technique for multi-area power system with SMSE unit, Electric
Power Components and Systems, Vol. 32, pp. 193-213.
Demiroren and E. Yesil, “Automatic generation control with fuzzy logic controllers in the power system including SMES units”,
Electrical Power & Energy Systems, Vol.26, No.4, pp. 291-305, May 2004.
Ghoshal S.P., June 2004, Optimizations of PID gains by particle swarm optimizations in fuzzy based automatic generation control,
Electrical Power System Research, Vol.2, pp.203-212.
Issarachai Ngamroo, 2005, Robust decentralized frequency stabilizers for SMES taking into consideration system uncertainties,
Electric Power System Research, Vol. 74, No. 2, pp. 281-292.
Kennady, J. and R.C. Eberhart, 1995, Particle swarm optimization. In: Proc. IEEE Int. Conf. on Neural Network, Perth, Australia,
pp: 1942-1948.
Kennady J., R.C. Eberhart and Y.Shi, “Swarm Intelligence”, Morgan Kaufmann, San Francisco, 2001.
Lim, K.Y. et al., 1996. Robust decentralized load frequency control of multi-area power system. IEE Proceedings-C, Vol. 143,
No. 5, pp. 377-386.
Nagpal M., A. Moshref, G.K Morision, P.Kundur, Jan 2001, Experience with Testing and modeling of gas turbine, Proceedings of
the IEEE/PES 2001 Winter Meeting, Columbus USA, pp. 652-656.
Pan, C.T. and C.M. Liaw, 1989. An adaptive control for power system LFC. IEEE Trans. PWRS, Vol. 4, No. 1, pp. 122-128.
Shayegi H., H.A. Shayanfar and O.P. Malik, 2007. Robust decentralized neural networks based LFC in a deregulated power
system. Electric Power System Research, Vol. 77, pp.241-251.
Shayeghi H. et al., 2007. Robust modified GA based multi-stage fuzzy LFC. Energy Conversion and Manag., Vol. 48, pp. 1656-
1670.
Shayeghi, H., H.A. Shayanfar, 2010. Application of PSO for fuzzy load frequency design with considering superconducting
magnetic energy storage, International Journal of TPE, Vol. 1, No. 2, pp. 24-33.
Shi Y. and R.C.Eberhart, 1998, Parameter selection in PSO”, Proceedings of 7th Annual Conference on Evolutionary Computation,
pp.591-601.
Soon Klat Yee, Jovica, F.Michael Hughcs, Feb 2008, Overview and comparative analysis of gas turbine model for system stability
studies, IEEE Transactions on Power Systems, Vol. 23, pp.108-118.
Tripathy S.C., R. Balasubramanian,P.S. Chandramohanan Nair, 1997, Adaptive automatic generation control with superconducting
magnetic energy storage in power systems, IEEE Trans. Energy Convers. Vol. 7, No. 3, pp. 434-441.
Wang Y., D.J. Hill and G. Guo, 1998. Robust decentralized control for multi-machine power system, IEEE Trans. on Circuits and
Systems: Fund. Theory and Applications., Vol. 45, No. 3, pp. 271-279.
Yamashita K. and H. Miagi, 1991. Multi variable self-tuning regulator for LFC system with interaction of voltage on load demand.
IEE Proceeding Conference: Theory and Applications. Vol. 138, No. 2, pp. 177-183.

Biographical Notes

R. Jayanthi received her B.E and M.E degrees from Faculty of Engineering and Technology, Annamalai University, Annamalai Nagar, Chidambaram, India in
1994 and 2007 respectively. Currently working as an Assistant Professor in Department of Electrical and Electronics Engineering since 2007. She is currently
working towards the Ph.D degree. Her research interest includes power system operation and control.

Dr. I. A. Chidambaram (1966) received Bachelor of Engineering in Electrical and Electronics Engineering (1987) Master of Engineering in Power System
Engineering (1992) and Ph.D in Electrical Engineering (2007) from Annamalai University, Annamalainagar. During 1988 - 1993 he was working as Lecturer in the
Department of Electrical Engineering, Annamalai University and from 2007 he is working as Professor in the Department of Electrical Engineering, Annamalai
University, and Annamalai Nagar. He is a member of ISTE and Indian Science Congress (ISC). His research interests are in power systems, electrical
measurements and control systems. (Electrical Measurements Laboratory, Department of Electrical Engineering, Annamalai University, Annamalai Nagar-608002,
TamilNadu, India, Tel: -91-04144-238501, Fax:-91-04144-238275) [email protected]/ [email protected]

C. Banusri is currently working as a Lecturer in Jaya College of Engineering, Avadi, Chennai, and received her B.E and M.E. degrees from Mailam Engineering
College affiliated to Madras University (2002) and Annamalai University (2010) respectively. Her areas of interest are power system modeling and control.

Received October 2010

Accepted April 2011
Final acceptance in revised form April 2011