Abstract—Design of a fixed parameter robust STATCOM controller for a multi-machine power system through an H-∞ based loop-shaping procedure is presented. The trial and error part of the graphical loop-shaping procedure has been eliminated by embedding a particle swarm optimization (PSO) technique in the design loop. Robust controllers were designed considering the detailed dynamics of the multi-machine system and results were compared with reduced order models. The robust strategy employing loop-shaping and PSO algorithms was observed to provide very good damping profile for a wide range of operation and for various disturbance conditions. Keywords—STATCOM, Robust control, Power system damping, Particle Swarm Optimization, Loop-shaping. I. INTRODUCTION HE static synchronous compensator (STATCOM) is a power electronics based synchronous voltage generator that generates a three-phase voltage from a dc capacitor. By controlling the magnitude of the STATCOM voltage the reactive power exchanges between the STATCOM and the transmission line and hence the amount of shunt compensation in the power system can be controlled [1]. In addition to reactive power exchange, a properly controlled STATCOM can also provide damping to a power system [2, 3]. A good number of recent literatures are available on modeling, operation and control fundamentals of the STATCOM [1, 4-5]. While most of the control designs are carried out with linearized models, nonlinear control strategies for STATCOM have also been reported recently [5]. STATCOM controls for stabilization have been attempted through complex Lyapunov procedures for simple power system models [6]. Applications of robust fuzzy logic and neural network based controls have also been reported [7, 8]. The controllers designed on the basis of linear theory are, generally, operating point dependent and hence are not robust in nature. A fixed parameter robust controller designed Manuscript received June 5, 2007. This work was supported by the electrical engineering department of King Fahd University of Petroleum and Minerals S.F. Faisal is with the Electrical Engineering Program, The Petroleum Institute, 2533, Abu Dhabi, UAE.( e-mail: sfaisal@pi.ac.ae). A.H.M.A. Rahim is with the department of Electrical Engineering, K.F. University of Petroleum and Minerals, Dhahran, Saudi Arabia (corresponding author:: P.O. Box# 349, Dhahran 31261, Saudi Arabia; phone: 0096638604986; fax:0096638603535; email:ahrahim@kfupm.edu.sa). J.M. Bakhashwain is with the department of Electrical Engineering, K.F. University of Petroleum and Minerals, Dhahran, Saudi Arabia. through graphical loop-shaping procedure was observed to provide good damping characteristics to a single machine power system [9]. However, application of such graphical techniques to multi-machine system is handicapped because of the higher order of the dynamics. This article presents an H-∞ based fixed parameter robust STATCOM controller design for a multi-machine power system. The realization of the robust controller for the high order multi-machine system through a graphical loop shaping procedure is simplified by embedding a particle swarm optimization (PSO) procedure in the design loop. Simulation results are presented comparing the design by original loop- shaping method as well as PSO based procedure. II. THE SYSTEM MODEL WITH STATCOM A 4-machine power system with STATCOMs located at the middle of the transmission lines connecting each generator to the rest of the grid is shown in Fig.1. The synchronous generator is represented by a two-axis model for the internal voltages and the swing equations; and its excitation system is assumed to be equipped with IEEE type-ST exciter model. The STATCOM is represented by a first order differential equation relating the STATCOM DC capacitor voltage and current. Fig. 1 A 4-machine power system with STATCOM A Robust STATCOM Controller for a Multi- Machine Power System Using Particle Swarm Optimization and Loop-Shaping S.F. Faisal, A.H.M.A. Rahim, J.M. Bakhashwain T International Journal of Electrical, Computer, and Systems Engineering 1;1 © www.waset.org Winter 2007 64