Anda di halaman 1dari 13

ISSN 1 746-7233, England, UK World Journal of Modelling and Simulation Vol. 6 (2010) No. 2, pp.

110-121

Modelling, simulation and optimal tuning of SSSC-based controller in a multi-machine power system
Sidhartha Panda Department of Electrical and Electronics Engineering, National Institute Of Science and Technology, Brahmapur 761008, India (Received February 13 2009, Revised July 4 2009, Accepted December 11 2009) Abstract. A systematic procedure for modeling, simulation and optimally tuning the parameters of a Static Synchronous Series Compensator (SSSC) controller in a multi-machine system, for power system stability enhancement is presented in this paper. For the simulation purpose, the model of multi-machine power system with SSSC controller is developed in MATLAB/SIMULINK using Sim Power System (SPS) blockset. The design problem of SSSC controller is formulated as an optimization problem and Real Coded Genetic Algorithm (RCGA) is employed to search for the optimal SSSC controller parameters. By minimizing a timedomain based objective function, in which the modal oscillations of the power system are involved; stability performance of the power system is improved. The results obtained from simulations validate the effectiveness of proposed modeling and tuning approach for power system stability improvement. The simulation results also show that the proposed SSSC controller is effective in damping a range of small disturbance conditions in the power system. Keywords: static synchronous series compensator, modelling and simulation, optimal tuning, real coded genetic algorithm, power system stability

Introduction

Recent development of power electronics introduces the use of Flexible AC transmission Systems (FACTS) controllers in power systems. FACTS controllers are capable of controlling the network condition in a very fast manner and this unique feature of FACTS can be exploited to improve the stability of a power system[18] . The detailed explanations about the FACTS controllers are well documented in the literature and can be found in [7, 12, 19]. Static synchronous series compensator (SSSC) is one of the important members of FACTS family which can be installed in series in the transmission lines. SSSC is very effective in controlling power ow in a transmission line with the capability to change its reactance characteristic from capacitive to inductive[6] . An auxiliary stabilizing signal can also be superimposed on the power ow control function of the SSSC so as to improve power system stability[20] . Mihalic and Papic[14] presented the basic differences between impedance based controllable series compensation and a power electronic converters with GTO thyristors based SSSC. For the time-domain simulation, the authors used a decoupled control structure with a predictive control loop based on the transformation of the three-phase system to the rotating reference frame. The design, modeling and control design of a 48-step inverter based SSSC suitable for subsynchronous resonance (SSR) analysis was presented by Kumar and Ghosh[10] . The equivalent circuit model of the SSSC was derived from a magnetic circuit containing 18 single-phase three winding transformers and six single-phase two winding transformers to add the output voltages of eight 3-phase inverters. Menniti et al.[13] proposed the use of SSSC to damp the transient frequency deviation in a deregulated electric power system. The authors proposed a method based on the application of the overlapping decomposition technique,

Corresponding author. E-mail address: panda sidhartha@rediffmail.com.

Published by World Academic Press, World Academic Union

World Journal of Modelling and Simulation, Vol. 6 (2010) No. 2, pp. 110-121

111

to design a decentralised control law of a SSSC device where a multi-area power system is decomposed into two decoupled subsystems. Ngamroo et al.[16] developed a robust decentralised frequency stabilisers design of SSSC by taking system uncertainties into consideration and proposed to use a SSSC in an interconnected power system which was subjected to load disturbances with changing frequency in the vicinity of the interarea oscillation mode. Jowder and Ooi[9] showed that addition of dielectric capacitors lowers the cost of series compensation by SSSC and reported based on HYPERSIM simulation that for an overall degree of capacitive compensation of 0.7 pu, the SSSC component is only about 1/3 of the capacitive MVAR and at the same time the transmissibility of the line is increased by a factor of 2.23 based on the transient stability limit. Jowder[8] investigated the inuence of the constant reactance and the constant quadrature voltage modes on the operation of a radial power system and showed that the constant reactance mode provides higher damping power, synchronizing power, and transient stability limit. Castro et al.[4] investigated the impacts of three different SSSC control modes on small-signal and transient stability of a power system, where the stability analysis and the design of the SSSC controllers were based on modal analysis, non-linear simulations, pole placement technique, and time and frequency response techniques and concluded that the usage of the SSSC in the constant impedance emulation mode is the most benecial strategy to improve both the small-signal and transient stability. Most of these proposals are based on small disturbance analysis that requires linearization of the system involved. However, linear methods cannot properly capture complex dynamics of the system, especially during major disturbances. This presents difculties for tuning the FACTS controllers in that the controllers tuned to provide desired performance at small signal condition do not guarantee acceptable performance in the event of major disturbances. Also, the performance of the controller under unbalanced faults can not be evaluated by using the linear single-phase models. In order to overcome the above shortcomings, this paper uses three-phase models of SSSC and power system components. A conventional lead-lag controller structure is preferred by the power system utilities because of the ease of on-line tuning and also lack of assurance of the stability by some adaptive or variable structure techniques. A number of conventional techniques have been reported in the literature pertaining to design problems of conventional power system stabilizers namely: the eigenvalue assignment, mathematical programming, gradient procedure for optimization and also the modern control theory. Unfortunately, the conventional techniques are time consuming as they are iterative and require heavy computation burden and slow convergence. In addition, the search process is susceptible to be trapped in local minima and the solution obtained may not be optimal[2] . The evolutionary methods constitute an approach to search for the optimum solutions via some form of directed random search process. A relevant characteristic of the evolutionary methods is that they search for solutions without previous problem knowledge. Recently, genetic algorithm (GA) appeared as a promising evolutionary technique for handling the optimization problems. GA as an optimization technique has advantage as it adapts to irregular search space unlike other conventional techniques. GA is becoming popular for solving the optimization problems in different elds of application, mainly because of its robustness in nding an optimal solution and ability to provide a near-optimal solution close to a global minimum. Unlike strict mathematical methods, the GA does not require the condition that the variables in the optimization problem be continuous and different; it only requires that the problem to be solved can be computed. GA employs search procedures based on the mechanics of natural selection and survival of the ttest. The GAs, which use a multiple-point instead of a single-point search and work with the coded structure of variables instead of the actual variables, require only the objective function, thereby making searching for a global optimum simple[5] . Therefore, in the present work GA is employed to search the optimal controller parameters. In this paper, a comprehensive assessment of the effects of SSSC-based damping controller for power system stability improvement in multi-machine power system has been carried out. First a simple lead-lag structure based controller for SSSC is proposed. Then, GA based optimal tuning algorithm is used to optimise the parameters of the SSSC controller. The design objective is to improve the stability of a multi-machine power system, subjected to severe disturbances. Simulation results show the advantages of using the modelling and tuning method when performing control and stability analysis in a power system involving a SSSC controller. The remainder of this paper is organized in ve major sections. A power system under study is presented in section 2. The proposed SSSC controller structure, problem formulation and optimization problem
WJMS email for subscription: info@wjms.org.uk

system stability improvement in multi-machine power system has been carried out. First a simple lead-lag structure based controller for SSSC is proposed. Then, GA based optimal tuning algorithm is used to optimise the parameters of the SSSC controller. The design objective is to improve the stability of a multi-machine power system, subjected to severe disturbances. Simulation results show the advantages of 112 using the modelling and tuning method when S. Panda: SSSC-based stability analysismulti-machine power system performing control and controller in a in a power system involving a SSSC controller. are explainedThesection 3. In section 4,is brief overview of realsections.genetic algorithm is given. Application in remainder of this paper a organized in five major coded A power system under study is of GA andpresented in section 2. The proposed SSSC controller section 5. Finally, in sectionandconclusions are given. simulation results is provided and discussed in structure, problem formulation 6 optimization problem are explained in section 3. In section 4, a brief overview of real coded genetic algorithm is given. Application of GA and simulation results is provided and discussed in section 5. Finally, in section 6 conclusions under 2 Power system are given.study

2. Power System Under Study To design and optimize the SSSC-based damping controller, a multi-machine system with SSSC, shown To design and optimize the SSSC-based damping controller, a multi-machine system 17]. The shown in Fig. 1, is considered. It is similar to the power systems used in references [15,with SSSC, system consists in Fig. 1, considered. It two subsystems and are used in references intertie. Following a disturbance, of three generatorsisdivided in tois similar to the power systemsconnected via an[16-17]. The system consists of three generators divided in to two subsystems and are connected via an intertie. Following a the two subsystems swing against each other resulting in instability. instability. To improve the stability To improve the stability the line is secdisturbance, the two subsystems swing against each other resulting in tionalized the lineSSSC is assumed a SSSC is assumed on the mid-point of the tieline. In the1Figure, G1,G3 and and a is sectionalized and on the mid-point of the tieline. In the Figure, G , G2 and G2 represent the generators; T represent /F3generators; T/F1 transformers and L1 , L2 and L3 represent the line sectionsline G3 /F1 T the represent the - T/F3 represent the transformers and L1, L2 and L3 represent the respectively.
sections respectively.

Bus2

L2

Bus4 Bus5 Bus6 L1 Bus1

G2
T/F2 Load2

L1

SSSC Bus3 L3 L1 L1 T/F1

G1

G3
T/F3 Load3 Load4

Load1

L1

Fig. 1. Multi-machine power system with SSSC

Fig. 1: Multi-machine power system with SSSC

3. The Proposed Approach 3.1. Structure of the SSSC-based damping controller


The structure of SSSC-based damping controller, to modulate the SSSC injected voltage Vq, is shown in

3
3.1

The proposed approach

Structure of the SSSC-based damping controller

The structure of SSSC-based damping controller, to modulate the SSSC injected voltage Vq , is shown in Fig. 2. The structure consists of a gain block with gain KS , a signal washout block and two-stage phase Fig. 2. The structure consists of a gain block with gain KS, a signal washout block and two-stage phase compensation block asblock as shown in 2. The The signal washoutblock servesas a a high-pass lter, with the time shown in Fig. Fig. 2. signal washout block serves as high-pass filter, with the compensation constant Ttime high enoughhigh allow signals associated with oscillations in input signal to pass unchanged. , constant TW, to enough to allow signals associated with oscillations in input signal to pass W From the viewpoint of thethe viewpoint of the washout function, the is notof TW is not critical and in the range of 1 to 20 unchanged. From washout function, the value of TW value critical and may be may be in the range of 1 to compensation The phase compensation T1S , (time constants T1S, 2S and T , T ) seconds[11] . The phase20 seconds. [18]. blocks (time constants blocks T2S and T3S , T4S )Tprovide3Sthe4Sappropriate phase-leadprovide the appropriate phase-lead characteristics to compensate for the phase lagthe output signals. In the Fig. characteristics to compensate for the phase lag between input and between input and the output signals. In the Fig. 2, Vqref represents the reference injected voltage as desired by the steady state 2, Vqref represents the reference injected voltage as desired by the steady state power ow control loop. The power flow control loop. The steady state power flow loop acts quite slowly in practice and hence, in the steady state power ow loopassumed to be constant practice and hence, intransient period. The desired value acts quite slowly in during large disturbance the present study Vqref is assumed to be present study Vqref is constant during large disturbance transient to the change in the SSSC injectedcompensation is is added to according of compensation is obtained according period. The desired value of voltage Vq which obtained to the change .in the SSSC injected voltage Vq which is added to Vqref . Vqref
Vq max Vqref + Vq Vq

KS

Input

sTW 1 + sTW

1 + sT1S 1 + sT2S

1 + sT3S 1 + sT4S

Vq
+

Gain block

Output
Vq min

Washout block

Two-stage lead-lag block

Vqref

Fig. 2. Structure of the SSSC-based damping controller


Fig. 2: Structure of the SSSC-based damping controller.

3.2. Problem formulation


The transfer function of the SSSC-based controller is: WJMS email for contribution: submit@wjms.org.uk

sTW U SSSC = K S 1 + sT W

1 + sT1S 1 + sT 2S

1 + sT3S 1 + sT 4S

(1)

World Journal of Modelling and Simulation, Vol. 6 (2010) No. 2, pp. 110-121

113

3.2

Problem formulation The transfer function of the SSSC-based controller is: USSSC = KS sTW 1 + sTW 1 + sT1S 1 + sT2S 1 + sT3S 1 + sT4S y, (1)

where, USSSC and y are the output and input signals of the SSSC-based controller respectively. In this structure, the washout time constants TW and the time constants T2S and T4S are usually prespecied[2, 3, 18] . In the present study, TW = 10s and T2S = T4S = 0.3s are used. The controller gain KS and the time constants T1S and T3S are to be determined. During steady state conditions Vq and Vqref are constant. During dynamic conditions the series injected voltage Vq is modulated to damp system oscillations. The effective Vq in dynamic conditions is given by: Vq = Vqref + Vq . 3.3 Optimization problem (2)

It is worth mentioning that the SSSC-based controller is designed to minimize the power system oscillations after a large disturbance so as to improve the power system stability. It is worth mentioning that the SSSC-based controller is designed to minimize the power system oscillations after a large disturbance so as to improve the power system stability. These oscillations are reected in the deviations in power angle, rotor speed and line power. Minimization of any one or all of the above deviations could be chosen as the objective. In the present study, an integral time absolute error of the speed signals corresponding to the local and inter-area modes of oscillations is taken as the objective function. The objective function is expressed as:
t=tsim

J=
t=0

L +

I t dt,

(3)

where, L and I are the speed deviations of inter-area and local modes of oscillations respectively and tsim is the time range of the simulation. In the present three-machine study, the local mode L is (2 3 ), and the inter-area mode I is [(2 1 ) + (3 1 )], where 1 , 2 and 3 are the speed deviations of machines, 1, 2 and 3 respectively. With the variation of the SSSC-based damping controller parameters, these speed deviations will also be changed. For objective function calculation, the time-domain simulation of the power system model is carried out for the simulation period. It is aimed to minimize this objective function in order to improve the system response in terms of the settling time and overshoots. The problem constraints are the SSSC controller parameter bounds. Therefore, the design problem can be formulated as the following optimization problem: Minimize J Subject to:
min max KS KS KS min T1S min T3S

(4)

(5) (6) (7)

T1S T3S

max T1S max T3S .

Tuning a controller parameter can be viewed as an optimization problem in multi-modal space as many settings of the controller could be yielding good performance. Traditional method of tuning doesnt guarantee optimal parameters and in most cases the tuned parameters needs improvement through trial and error. In GA based method, the tuning process is associated with an optimality concept through the dened objective function and the time domain simulation. The designer has the freedom to explicitly specify the required performance objectives in terms of time domain bounds on the closed loop responses. Hence the GA method yields optimal parameters and the method is free from the curse of local optimality. In view of the above, the proposed approach employs GA to solve this optimization problem and search for optimal set of SSSC-based damping controller parameters from throughout the world. A brief overview of genetic algorithm optimization technique is described in the next section.
WJMS email for subscription: info@wjms.org.uk

114

S. Panda: SSSC-based controller in a multi-machine power system

Overview of genetic algorithm

Genetic algorithm (GA) has been used to solve difcult engineering problems that are complex and difcult to solve by conventional optimization methods. GA maintains and manipulates a population of solutions and implements a survival of the ttest strategy in their search for better solutions. The ttest individuals of any population tend to reproduce and survive to the next generation thus improving successive generations. The inferior individuals can also survive and reproduce. Implementation of GA requires the determination of six fundamental issues: chromosome representation, selection function, the genetic operators, initialization, termination and evaluation function. Brief descriptions about these issues are provided in the following sections. 4.1 Chromosome representation

Chromosome representation scheme determines how the problem is structured in the GA and also determines the genetic operators that are used. Each individual or chromosome is made up of a sequence of genes. Various types of representations of an individual or chromosome are: binary digits, oating point numbers, integers, real values, matrices, etc. Generally natural representations are more efcient and produce better solutions. Real-coded representation is more efcient in terms of CPU time and offers higher precision with more consistent results. 4.2 Selection function

To produce successive generations, selection of individuals plays a very signicant role in a genetic algorithm. The selection function determines which of the individuals will survive and move on to the next generation. A probabilistic selection is performed based upon the individuals tness such that the superior individuals have more chance of being selected. There are several schemes for the selection process: roulette wheel selection and its extensions, scaling techniques, tournament, normal geometric, elitist models and ranking methods. The selection approach assigns a probability of selection Pj to each individuals based on its tness value. In the present study, normalized geometric selection function has been used. In normalized geometric ranking, the probability of selecting an individual Pi is dened as: Pi = q (1 q)r1 , q = q , 1 (1 q)P (8)

where, q = probability of selecting the best individual, r = rank of the individual (with best equals 1), p = population size. 4.3 Genetic operators

The basic search mechanism of the GA is provided by the genetic operators. There are two basic types of operators: crossover and mutation. These operators are used to produce new solutions based on existing solutions in the population. Crossover takes two individuals to be parents and produces two new individuals while mutation alters one individual to produce a single new solution. The following genetic operators are usually employed: simple crossover, arithmetic crossover and heuristic crossover as crossover operator and uniform mutation, non-uniform mutation, multi-non-uniform mutation, boundary mutation as mutation operator. Arithmetic crossover and non-uniform mutation are employed in the present study as genetic operators. Crossover generates a random number r from a uniform distribution from 1 to m and creates two new individuals by using equations: xi = xi , if i < r yi otherwise , yi = yi , if i < r xi otherwise . (9)

WJMS email for contribution: submit@wjms.org.uk

World Journal of Modelling and Simulation, Vol. 6 (2010) No. 2, pp. 110-121

115

Arithmetic crossover produces two complimentary linear combinations of the parents, where r = U (0, 1): X = rX + (1 r)Y , Y = rY + (1 r)X. (10)

Non-uniform mutation randomly selects one variable j and sets it equal to an non-uniform random number. xi + (bi xi ) f (G) if r1 < 0.5, xi + (xi + ai ) f (G) if r1 0.5, xi = , xi , otherwise where,
Y ' = r Y + (1 r ) X

(11)

(13)

, (12) f (G) = r 1 xi + (bi xi ) f (G ) if 2r1 < 0.5, Gmax (14) xi ' = xi + ( xi + a i ) f (G ) if r1 0.5, r1 , r2 =uniform random nos. xi , 0 to 1, G =currentgeneration, Gmax =maximum no. of generations, between otherwise
where, b =shape parameter.

Non-uniform mutation randomly selects one variable j and sets it equal to an non-uniform random b number. G

f (G ) = (r2 (1

G G max

)) b

(15)

4.4

Initialization, termination and evaluation function 1. r1, r2 = uniform random nos. between 0 to
G = current generation. G = maximum no. of generations. population maxshape parameter. the genetic is needed to start b=

An initial algorithm procedure. The initial population can be randomly generated or can be taken from other methods. The GA4.4. Initialization, termination and evaluation function moves from generation to generation until a stopping criterion is met. The stopping criterion An initial population is needed to startpopulationalgorithm procedure. The initial population can be in the best could be maximum number of generations, the genetic convergence criteria, lack of improvement randomly generated or can be taken from other methods. solution over a specied number generation to generation until a stopping criterion is met. The stopping criterion of generations or target value for the objective function. The GA moves from Evaluation functions or number of generations, population convergence criteria, lack ofaimprovement inthe function can could be maximum objective functions of many forms can be used in GA so that the best solution over a specified number set. The computational owchart of the GA map the population into a partially orderedof generations or target value for the objective function. optimization process employed in theEvaluation study is given in Fig. 3. of many forms can be used in a GA so that the function present functions or objective functions
can map the population into a partially ordered set. The computational flowchart of the GA optimization process employed in the present study is given in Fig. 3.
Start

Specify the parameters for GA

Generate initial population Gen.=1 Time-domain simulation of power system model Find the fitness of each individual in the current population

Gen.=Gen.+1

Yes Gen. > Max. Gen.? No Apply GA operators: selection,crossover and mutation Stop

Fig. 3. Flowchart of genetic algorithm

Fig. 3: Flowchart of genetic algorithm.

WJMS email for subscription: info@wjms.org.uk

116

PWM technique to synthesize a sinusoidal waveform from a DC voltage with a typical chopping frequency of a few kilohertz. Harmonics are cancelled by connecting lters at the AC side of the VSC. This type of VSC uses a xed DC voltage. The converter voltage is varied by changing the modulation index of the PWM modulator. A DC link nominal voltage of 40 kV and DC link equivalent capacitance of 375 F is used in the present study[1] . S. Panda: SSSC-based controller in a multi-machine power system

Fig. 4. MATLAB/SIMULINK model of example power system

Fig. 4. MATLAB/SIMULINK model of example power system

ResultsApplication of GA optimization technique and discussions 5.1


For the purpose of optimization of Eq. (3), RCGA is employed. For the implementation of GA normal

The Sim Power Systems (SPS), a product of MATLAB function based onin the present geometric all simfamily, is used the normalized study for geometric selection is employed which is a ranking selection ulations and SSSC-based damping controllerparents and .performs an modern design tool line formed bythe users design[1] SPS is a interpolation along the that allows distribution. Arithmetic crossover takes two to rapidly two parents. Non uniform mutation changes one of the parameters ofSPSparent basedin the Simulink environthe and easily build models and simulate the power systems. the operates on a non-uniform probability main library, Gaussian distribution three-phase narrows to typical power equipments such ment. The SPSs distribution. Thispowerlib, containsstarts wide, andmodels of a point distribution as the current generation approaches the maximum generation. as machines, governors, excitation systems, transformers, lines and FACTS devices. The Powergui block is The objective function is evaluated necessary for simulation of any Simulink for each individual by Sim Power Systems blocks. It provides useful model containing simulating the example power system, considering a severe disturbance. For objective function calculation, a 3-phase short-circuit fault in one of the graphical user transmission lines is considered. For different problems, it is possiblelibrary also contains the Powergui interface (GUI) tools for the analysis of SPS models. The that the same parameters for parallel block that opens give thefor the steady-state analysis of electrical circuits. It situation. One more important GA do not a GUI best solution and so these can be changed according to the performs load ows and initializes the three-phase networks optimal solution more or less is the range for unknowns. For the starts in execution of point that affects the containing three-phase machines so that the simulation very rst steady state. The phasor solution method is mainly used to study electromechanical oscillations of power systems conWJMS email for contribution: submit@wjms.org.uk sisting of large generators and motors. In a power system stability study, the fast oscillation modes resulting from the interaction of linear R, L, C elements and distributed parameter lines are of no interest. These oscillation modes, which are usually located above the fundamental frequency of 50 Hz or 60 Hz, do not interfere with the slow machine modes and regulator time constants. In the phasor solution method, these fast modes are ignored by replacing the networks differential equations by a set of algebraic equations. The state-space model of the network is therefore replaced by a transfer function evaluated at the fundamental frequency and relating inputs (current injected by machines into the network) and outputs (voltages at machine terminals). The phasor solution method uses a reduced state-space model consisting of slow states of machines, turbines, and regulators, thus dramatically reducing the required simulation time. Continuous variable-step solvers are very efcient in solving this type of problem. Recommended solvers are ode15s or ode23tb with a maximum time step of one cycle of the fundamental frequency (1/60 s or 1/50 s). In order to optimally tune the parameters of the SSSC-based damping controller, as well as to assess its performance and robustness under wide range of operating conditions with various fault disturbances and fault
WJMS email for contribution: submit@wjms.org.uk

World Journal of Modelling and Simulation, Vol. 6 (2010) No. 2, pp. 110-121

117

clearing sequences, the test system depicted in Fig. 1 is considered for analysis. The MATLAB/SIMULINK model of the example power system is developed using SPS blockset as shown in Fig. 4. The system consists of three hydraulic generating units divided into two subsystems. The ratings of the generators are taken as 1400 MVA each (G2 and G3 ) in one subsystem and 4200 MVA (G1 ) in the other subsystem. The generators are represented by a sixth-order model and are equipped with Hydraulic Turbine & Governor (HTG) and Excitation systems. The HTG represents a nonlinear hydraulic turbine model, a PID governor system, and a servomotor. The excitation system consists of a voltage regulator and DC exciter, without the exciters saturation function. The generators with output voltages of 13.8KV are connected to an intertie through 3-phase step up transformers. The machines are equipped with Hydraulic Turbine and Governor (HTG) and Excitation system. These components are included in Reg M1, Reg M2 and Reg M3 subsystem blocks respectively. These blocks are available in the SPS library powerlib/Machines[1] . The variation of injected voltage is performed by means of a voltage sourced converter (VSC). The VSC uses forced-commutated power electronic devices (e.g. GTOs, IGBTs or IGCTs) to synthesize a voltage from a DC voltage source. A capacitor connected on the DC side of the VSC acts as a DC voltage source. VSC using IGBT-based PWM inverters is used in the present study to model the SSSC. This type of inverter uses PWM technique to synthesize a sinusoidal waveform from a DC voltage with a typical chopping frequency of a few kilohertz. Harmonics are cancelled by connecting lters at the AC side of the VSC. This type of VSC uses a xed DC voltage. The converter voltage is varied by changing the modulation index of the PWM modulator. A DC link nominal voltage of 40 kV and DC link equivalent capacitance of 375 F is used in the present study[1] . 5.1 Application of GA optimization technique

For the purpose of optimization of Eq. (3), RCGA is employed. For the implementation of GA normal geometric selection is employed which is a ranking selection function based on the normalized geometric distribution. Arithmetic crossover takes two parents and performs an interpolation along the line formed by the two parents. Non uniform mutation changes one of the parameters of the parent based on a non-uniform probability distribution. This Gaussian distribution starts wide, and narrows to a point distribution as the current generation approaches the maximum generation. The objective function is evaluated for each individual
Table 1. Bounds of unknown variables and optimized parameters by RCGA Parameters Minimum range Maximum range Optimised parameters obtained by GA Gain KS 10 500 238.031 Time constants T1S T3D 0.01 0.01 0.7 0.7 0.0761 0.4272

by simulating the example power system, considering a severe disturbance. For objective function calculation, a 3-phase short-circuit fault in one of the parallel transmission lines is considered. For different problems, it is possible that the same parameters for GA do not give the best solution and so these can be changed according to the situation. One more important point that affects the optimal solution more or less is the range for unknowns. For the very rst execution of the program, more wide solution space can be given and after getting the solution one can shorten the solution space nearer to the values obtained in the previous iteration. Optimization was performed with the total number of generations set to 100. The convergence rate of objective function J with the number of generations is shown in Fig. 5. Bounds for unknown parameters of gain and time constants used in the present study and the obtained parameters of SSSC controller by the GA run are given in are shown in Tab. 1. 5.2 Simulation results

In order to assess the effectiveness and robustness of the proposed controller, simulation studies are carried out under different contingencies. The response without controller is shown with dotted lines with
WJMS email for subscription: info@wjms.org.uk

one can shorten the solution space nearer to the values obtained in the previous iteration. Optimization was performed with the total number of generations set to 100. The convergence rate of objective function J with the number of generations is shown in Fig. 5. Bounds for unknown parameters of gain and time constants used in the present study and the obtained parameters of SSSC controller by the GA run 118 are given in are shown in Table I. S. Panda: SSSC-based controller in a multi-machine power system
x 10 5.55 5.5
-3

Convergence of fitness

5.45 5.4 5.35 5.3 5.25 5.2 5.15 0 10 20 30 40 50 60 70 80 90 100

Generations

Fig. 5: Convergence of fitness function. Fig. 5. Convergence of tness function


Table I: Bounds of unknown variables and optimized parameters by RCGA

legend NC and responseParameters with GA optimized SSSC-based damping controller is shown with solid lines with Gain Time constants T1S T3S legend SSSC. The following cases are considered: KS
Minimum range Maximum range Case 1: three-phase fault disturbance Optimised parameters 10 500 238.031 0.01 0.7 0.0761 0.01 0.7 0.4272

obtained by GA A 3-cycle, 3-phase fault is applied at one of the line sections between bus1 and bus 6 near bus 6 at t = 1s. The fault is cleared by opening the faulty line and the line is reclosed after 3-cycles. Fig. 6 Fig. 8 show the variations of the inter-area and local mode of oscillation against time. From these Figs. it can be seen that, 5.2. modes of oscillations inter-area Simulation results are highly oscillatory in the absence of SSSC-based damping controller and the In order tosignicantly effectiveness and robustness stabilityproposed controller, oscillations. studies are controller assess the improves the power system of the by damping these simulation Further, the proposed controller is also effective in suppressing the local mode controller is shown with dotted lines with carried out under different contingencies. The response without of oscillations. The SSSC injected voltage Vq for the NC and response with GA optimized SSSC-based damping is effectivelyshown with solid lines legend above contingency is shown in Fig. 9 which shows that Vq controller is modulated to suppress power system oscillations. following cases are considered: with legend SSSC. The

Case 2: small disturbance disturbance Case 1: Three-phase fault


AIn order to examine the effectiveness ofof the line sections between bus1 and disturbance, the load=at bus 3-cycle, 3-phase fault is applied at one the proposed controller under small bus 6 near bus 6 at t 1 s. The fault is cleared by opening ms. faulty line and the line is the inter-area and local modes 6-8oscillations 4 is disconnected at t = 0.1s for 50 the Fig. 10 and Fig. 11 show reclosed after 3-cycles. Figs. of show the variations from inter-area and local mode SSSC-based against time. From damps the can be seen that, against time, of the which it is clear that the of oscillation damping controllerthese Figs. itmodal oscillations inter-area modes of oscillations are effectively even for small disturbance. highly oscillatory in the absence of SSSC-based damping controller and the controller significantly improves the power system stability by damping these oscillations. Case 3: unbalanced fault controller is also effective in suppressing the local mode of oscillations. The SSSC Further, the proposed disturbance

The effectiveness of the proposed controller to unbalanced faults is also examined by applying self clearing type unsymmetrical faults (namely double line-to-ground, line-to-line and single line-to-ground) of 3-cycle duration, at bus 1. The inter-area and local modes of oscillations against time are shown in Fig. 12 and Fig. 13 respectively. In these Figs., the uncontrolled system response for least severe single line-to-ground fault is also shown with dotted lines. It is clear from the simulation results that the modal oscillations are poorly damped in uncontrolled case even for the least severe fault and the SSSC-based damping controller effectively stabilizes the modal oscillations under various unbalanced fault conditions.

Conclusions

This paper presents a systematic procedure for modeling, simulation and optimal tuning of SSSC controller in a multi- machine system for enhancing power system stability. For the SSSC controller design problem, a parameter-constrained, time-domain based, objective function, is developed to improve the performance of power system subjected to a disturbance. Then, real coded genetic algorithm employed to search for the
WJMS email for contribution: submit@wjms.org.uk

modulated to suppress powerabove contingency is shown in Fig. 9 which shows that Vq is effectively injected voltage Vq for the system oscillations. injected voltage Vq forpower system oscillations. shown in Fig. 9 which shows that Vq is effectively modulated to suppress the above contingency is modulated to suppress power system oscillations. -4
World Journal of Modelling and Simulation, Vol. 6 (2010) No. 2, pp. 110-121 -4
10 x 10 10x 10 10 5 5 5 0 0 0 -5 -5 -5 0 0 1 1 2 2 3 3 4 5 6 4 Time5(sec) 6 7 7 8 8 9 9 10 10
-4

x 10

119
NC SSSC NC SSSC NC SSSC

Fig. 6: 0 Inter-area mode of oscillation ( 2 1 ) for73-phase fault disturbance 1 2 3 4 Time (sec) 6 5 8 9 10 Time (sec) Fig. 6: Inter-area mode of oscillation ( 2 1 ) for 3-phase fault disturbance Fig.Fig.Inter-area mode ofof oscillation((2 1 ) for 3-phase fault disturbance 6: 6. 10-4 oscillation 2 1 ) for 3-phase fault disturbance x Inter-area mode
10 x 10 10 -4 x 10 10 5
-4

2 - 1- -(pu) (pu) (pu)

2 1

NC SSSC NC SSSC NC SSSC

3 - 1- -(pu) (pu) (pu)

5 5 0 0 0 -5 -5 -5 0 0 1 1 2 2 3 3 4 5 6 4 Time5(sec) 6 7 7 8 8

3 1

9 9

10 10

Fig. 7: 0Inter-area mode of 3 oscillation ( (sec)1 ) for73-phase fault disturbance 1 2 4 Time 3 6 5 8 9 10 Time (sec) Fig. 7: Inter-area mode of oscillation ( 3 1 ) for 3-phase fault disturbance Fig. 7. Inter-area mode of oscillation (3 1 ) for 3-phase fault disturbance -4 Fig. 7: Inter-area mode of oscillation ( 3 1 ) for 3-phase fault disturbance x 10
x 10 4 -4 x 10 4
-4

NC SSSC NC SSSC NC SSSC

2 - 3- -(pu) (pu) (pu)

42 2 20 0 -2 0 -2 -4 -2 -4 0 1 2 3 4 5 6 7 8

2 3

10

Fig. 8: Local mode of oscillation Time (sec) 3 ) for 3-phase fault disturbance ( 2 Fig. 8: Local mode of oscillation ( 2 3 ) for 3-phase fault disturbance

-4 0 1 2 3 4 Time5(sec) 6 7 8 9 10 Fig. 8: 8. Local mode of oscillation (( 2 36) for 3-phase8fault disturbance 0 Local mode of oscillation Time ) for 3-phase fault disturbance 1 2 3 4 5 (sec) 7 9 10 Fig.
2 3

optimal SSSC controller parameters. The controller is tested on example power system subjected to various types of disturbances. The simulation results show that, the genetically tuned SSSC controller improves the stability performance of the power system and power system oscillations are effectively damped out under severe disturbance conditions. Further it is observed that the proposed SSSC controller is effective in damping the modal oscillations resulting from unbalanced fault and small disturbance conditions. It can be concluded
WJMS email for subscription: info@wjms.org.uk

0.15

Vq Vq (pu) (pu)

0.1 0.1 0.05 0.05 0 0


0.2

120

S. Panda: SSSC-based controller in a multi-machine power system


NC SSSC

Case 2: Small disturbance Case 2: Small disturbance 0

Fig. 9: SSSC injected voltage ( VqTime (sec) for 3-phase fault disturbance ) variation 0.05 Fig. 9: SSSC injected voltage ( Vq ) variation for 3-phase fault disturbance

In order to examine the effectiveness of the proposed controller under small disturbance, the load at bus 4 is -0.05 In order to at t = 0.1 s for 50 ms. Figs. 10 and 11 show the inter-area and local disturbance, the load at bus 4 is disconnectedexamine the effectiveness of the proposed controller under small modes of oscillations against time, disconnected t clear that 50 SSSC-based damping controller 4damps the modal oscillations effectively even for 1 2 3 5 6 from which itatis = 0.1 s for the0 ms. Figs. 10 and 11 show the inter-area and local modes of7oscillations against time, Time (sec) damps the modal oscillations effectively even for from disturbance. small which it is clear that the SSSC-based damping controller Fig. 9: SSSC injected voltage ( Vq ) variation for 3-phase fault disturbance small disturbance. Fig. 9. SSSC injected voltage (Vq ) variation for 3-phase fault disturbance -4 x 10 Case 2: Small disturbance -4
In order to examine the effectiveness of the proposed controller under small disturbance, the load at bus 4 is 3 disconnected at t = 0.1 2 for 50 ms. Figs. 10 and 11 show the inter-area and local modes of oscillations against time, s 2 from which it is clear 1 that the SSSC-based damping controller damps the modal oscillations effectively even for small disturbance. 1
2 - 1 2 - 1 (pu)(pu)
0 0 -1 -1 -2
x 10 3 2 1 0 -1
-4

Vq (pu)

-0.05 -0.05 0.15 0 0 0.1

1 1

2 2

3 4 Time (sec)4 3

5 5

6 6

7 7

3 x 10

-30

2 - 1 (pu)

-2 -3

NC SSSC NC 1SSSC 1 2 2 3 4 Time (sec)4 3 5 5 6 6 7 7

Fig. 10: Inter-area mode of 2 1 ) for small disturbance Fig. 10: Inter-area mode of oscillation (2 2 11 ) for small disturbance Fig. 10.-2 Inter-area mode of oscillation ( ) for small disturbance
Time ( oscillation(sec)
8 8 6 6 x 10
-5 -5

NC

-3 0

SSSC 1 2 3 4 5 6

x 10

Time (sec)

4 SSSC Fig. 10: Inter-area mode of oscillation ( 2 1 ) for small disturbance

NC 7 SSSC NC

2 - 3 (pu) 2 - 3 (pu)

4 2 2 0 0 -2 -2 -4 -4 -6 -6 -8 -8
8 6 4 2 0 -2 -4 x 10
-5

NC SSSC

2 - 3 (pu)
0

1 1

2 2

3 3

5 5

6 6

7 7

8 8

Time 4 (sec) Time (sec)

Fig. 11:-6Local mode of oscillation ( 2 3 ) for small disturbance Fig. 11. Local mode of oscillation (2 3 ) for small disturbance Fig. 11: Local mode of oscillation ( 2 3 ) for small disturbance
-8 0 1 2 3 4 5 6 7 8

that, the local and inter-area modes of oscillations ofTime (sec) system can be effectively damped for various power Fig. 11: Local mode of oscillation ( 2 3 ) for small disturbance disturbances by using the proposed SSSC controller.

References
[1] Sim Power Systems 4.3, Users guide. Http://www.mathworks.com/products/simpower/. [2] Y. Abdel-Magid, M. Abido. Robust coordinated design of excitation and TCSC-based stabilizers using genetic algorithms. Electrical Power and Energy Systems, 2004, 69(2-3): 129141.
WJMS email for contribution: submit@wjms.org.uk

World Journal of Modelling and Simulation, Vol. 6 (2010) No. 2, pp. 110-121
3 3 2 2 x 10-4 x 10
-4

Figs., the uncontrolled system response for least severe single line-to-ground fault is also shown with dotted lines. It Figs., the uncontrolled system response for least severe single line-to-ground fault is also shown with dotted lines. It is clear from the simulation results that the modal oscillations are poorly damped in uncontrolled case even for the is clear from the simulation results that the modal oscillations are poorly damped in uncontrolled case even for the least severe fault and the SSSC-based damping controller effectively stabilizes the modal oscillations under various least severe fault and the SSSC-based damping controller effectively stabilizes the modal oscillations under various unbalanced fault conditions. unbalanced fault conditions.

121

22--11(pu) (pu)

1 1

2 2

1: L-L-G fault 1: L-L-G fault 2: L-L fault 2: L-L fault 3: L-G fault 3: L-G fault 4: L-G fault uncontrolled 4: L-G fault uncontrolled

0 0 -1 -1

3 3 1 1

4 4

-2 -2

0 0

0.5 0.5

1 1

1.5 1.5

2 2

2.5 2.5

3 3

3.5 3.5

4 4

4.5 4.5

5 5

Fig. 12: Inter-area mode of oscillation ( ) for small disturbance Fig. 12: Inter-area mode of oscillation ( 2 1 ) for small disturbance 2 1 Fig. 12. Inter-area mode of oscillation (2 1 ) for small disturbance
x 10-4 x 10 2 2 1: L-L-G fault 1: L-L-G fault 2: L-L fault 2: L-L fault 3: L-G fault 3: L-G fault 4: L-G fault uncontrolled 4: L-G fault uncontrolled
-4

Time (sec) Time (sec)

1 1 2 2 4 4

22 --33 (pu) (pu)

1 1

0 0

-1 -1

3 3

-2 -2

0 0

1 1

2 2

3 3

4 4

5 5

6 6

Fig. 13: Local mode of oscillation ( 2 3 ) for small disturbance Fig. 13: Local mode of oscillation ( 3 ) for small disturbance Fig. 13. Local mode of oscillation ( 22 3 ) for small disturbance

Time (sec) Time (sec)

6. Conclusions
This paper presents a systematic procedure for modeling, simulation and optimal tuning of SSSC controller in a [3]This Abido.presents a and assessment of STATCOM-basedsimulationstabilizers for tuning of SSSC controller in a M. paper Analysis systematic procedure for modeling, damping and optimal power system stability enhancemulti- machine system for enhancing power system stability. For the SSSC controller design problem, a parametermulti- machine system for enhancing power system stability. For the SSSC controller design problem, a parameterment. Electric Power Systems Research, 2005, 73: 177185. constrained, time-domain based, objective function, is developed to improve the performance of power system power system. [4]constrained, H. Ayres, et al. Impacts of the SSSC control developed small-signal transient stability of a powersystem M. Castro, time-domain based, objective function, is modes on to improve the performance Electric to a disturbance. Then, real coded genetic algorithm employed to search for the optimal SSSC controller Power System Research, 2007, 77: 19. subjected to a disturbance. Then, real coded genetic algorithm employed to search for the optimal SSSC controller subjected [5]parameters. The controller is tested on example power systemMachine Learning. M. types of disturbances. The D. Goldberg. Genetic Algorithms in Search, Optimization, and subjected to various A: Addison-Wesley, Boston, parameters. The controller is tested on example power system subjected to various types of disturbances. The 1989. simulation results show that, the genetically tuned SSSC controller improves the stability performance of the power simulation results show that, the genetically tuned SSSC controller improves the stability performance of the power [6] L. Gyugyi, C. Schauder, K. Sen. Static synchronous series compensator: a solid-state approach to the series compensation of transmission lines. IEEE Transactions on Power Delivery, 1997, 12(1): 406417. [7] N. Hingoran, L. Gyugyi. Understanding FACTS: Concepts and Technology of Flexible AC Transmission Systems. IEEE Press, New York, 2000. [8] F. Jowder. Inuence of mode of operation of the SSSC on the small disturbance and transient stability of a radial power system. IEEE Transactions on Power Systems, 2005, 20(2): 935942. [9] F. Jowder, B. Ooi. Series compensation of radial power system by a combination of sssc and dielectric capacitors. IEEE Transactions on Power Delivery, 2005, 20(1): 458465. [10] L. S. Kumar, A. Ghosh. Static synchronous series compensator-design, control and application. Electric Power System Research, 1999, 49: 139148. [11] P. Kundur. Power System Stability and Control. McGraw-Hill, New York, 1994. [12] R. Mathur, R. Verma. Thyristor-based FACTS Controllers for Electrical Transmission Systems. IEEE press, Piscataway, 2002. [13] D. Menniti, A. Pinnarelli, et al. Using a FACTS device controlled by a decentralized control law to damp the transient frequency deviation in a deregulated electric power system. Electric Power System Research, 2004, 72: 289298. [14] R. Mihalic, I. Papic. Static synchronous series compensator-a mean for dynamic power ow control in electric power systems. Electric Power System Research, 1998, 45: 6572.
WJMS email for subscription: info@wjms.org.uk

122

S. Panda: SSSC-based controller in a multi-machine power system

[15] S. Mishra, P. Dash, et al. Genetically optimized neuro-fuzzy IPFC for damping modal oscillations of power systems. IEEE Transactions on Power Systems, 2002, 17(4): 11401147. [16] I. Ngamroo, J. Tippayachai, S. Dechanupaprittha. Robust decentralised frequency stabilizers design of static synchronous series compensators by taking system uncertainties into consideration. International Journal of Electrical Power and Energy Systems , 2006, 28: 513524. [17] M. Noroozian, G. Anderson, K. Tomsovic. Robust near-optimal control of power system oscillation with fuzzy logic. IEEE Transactions on Power Delivery , 1996, 11(1): 393400. [18] S. Panda, N. Padhy, R. Patel. Modelling, simulation and optimal tuning of TCSC controller. International Journal of Simulation Modelling, 2007, 6(1): 3748. [19] Y. Song, T. Johns. Flexible AC Transmission Systems (FACTS). IEE, London, 2000. [20] H. Wang. Static synchronous series compensator to damp power system oscillations. Electric Power System Research, 2000, 54: 113119.

Appendix
A complete list of parameters used appears in the default options of Sim Power Systems in the Users Manual[1] . All data are in pu unless specied otherwise. SB1 = 4200MVA, SB2 = SB3 = 2100WVA, H = 3.7s, VB = 13.8kV, f = 60Hz, RS = 2.8544e3 , Xd = 1.305, Xd = 0.296, Xd = 0.252, Generators Xq = 0.474, Xq = 0.243, Xq = 0.18, Td = 1.01 s, Td = 0.053 s, Tqo = 0.1 s, Pe1 = 0.3048, Pe2 = 0.2095, Pe3 = 0.1445 Load1 = 15000MW + 1500MVAR, Load2 = Load3 = 25MW, Loads Load4 = 250MW SBT1 = 2100MVA, SBT2 = SBT3 = 2100MVA, 13.8/500kV , 60Hz, R1 = R2 = 0.002, L1 = 0, L2 = 0.12, D1/Yg connection, Transformers Rm = 500, Lm = 500 3-Ph, 60Hz, L1 = 175km, L2 = 50km, L3 = 100km, Transmission lines R1 = 0.02546/km, R0 = 0.3864/km, L1 = 0.9337e3 H/km, L0 = 4.1264e3 H/km, C1 = 12.74e9 F/km, C0 = 7.751e9 F/km Ka = 3.33, Ta = 0.07, Gmin = 0.01, Gmax = 0.97518, Hydraulic Turbine and V gmin = 0.1pu/s, Vgmax = 0.1pu/s, Rp = 0.05, Kp = 1.163, Governor Ki = 0.105, Kd = 0, Td = 0.01s, = 0, Tw = 2.67 s TLP = 0.02s, Ka = 200, Ta = 0.001s, Ke = 1, Te = 0, Tb = 0, Excitation System Tc = 0, Kf = 0.001, Tf = 0.1s, Efmin = 0, Efmax = 7, Kp = 0 Snom = 100MVA, Vnom = 500kV, f = 60Hz, Vqmax = 0.2, Max rate of change of Vqref = 3/s, Rcnv = 0.00533, Lcnv = 0.16, SSSC VDC = 40kV, CDC = 375e6 F, KP IVR = 0.00375, KI IVR = 0.1875, KP VdcR = 0.1e3 , KP VdcR 20e3

WJMS email for contribution: submit@wjms.org.uk