Dynamic reactive power optimization using mathematical morphology and genetic algorithm

A new approach of dynamic reactive power optimization is presented in this paper, which is based on mathematical morphology and genetic algorithm. Due to the difficulty of controlling the compensatorpsilas operation-time-number, a mathematical morphology filter is used to transfer the problem into filtering alternative images consisting of alleles on chromosomes. At the same time, some improvements are made in crossover and mutation for accelerating the speed of genetic algorithm, in which the genetic character is introduced to avoid breaking the excellent combination of genes and according to the correction of particle velocity derived from particle swarm optimization, an evolutional mutation based on excellent genetic character is presented. The practice in a distribution network proves that the algorithm presented in this paper is right and effective.