Application of particle swarm optimization technique for reactive power optimization problem

Abstract

Reactive power optimization is an important function both in planning and day-to-day operation of power systems. Reactive power planning utilizes the existing reactive power sources judiciously while planning suitable location and size of VAR compensation to be installed in a system to meet the requirements of future load increase to keep reasonable voltage profile under normal condition and contingency states. RPP considers minimization of the allocation and operational cost as objectives while satisfying the constraints that define satisfactory operation of the system. These objectives can be achieved by proper adjustment of reactive power control variables, such as reactive power outputs of generators, tap ratios of transformers and outputs of shunt capacitors/reactors. Evolutionary computation techniques such as genetic algorithm (GA), Evolutionary Programming (EP), and Evolutionary Strategy (ES) have been proposed to solve the reactive power optimization problem. This thesis proposes Particle Swarm Optimization (PSO) one of the recently introduced population based optimization technique to solve the reactive power planning problem. In this work, the voltage stability level of the system is included as an additional objective of the RPP problem and this multi objective reactive power planning problem is solved using vector evaluated particle swarm optimization (VEPSO) technique. VEPSO is a multi objective variant of particle swarm optimization. In VEPSO, the population is divided into number of sub populations based on the number of objective functions. This thesis proposes a hybrid particle swarm optimization algorithm for reactive power optimization in electricity market. In the proposed HPSO, to avoid the possibility of trapping in local optimum, genetic operators namely crossover and mutation are applied to the PSO algorithm. Hybrid particle swarm algorithms gives better results than the simple genetic algorithm and simple particle swarm optimization algorithm when applying to reactive power optimization