Improved particle swarm optimization algorithm and its application in power electronic controller

Abstract

The traditional particle swarm optimization (PSO) algorithm has been used to optimize inverter controller parameters. However, there are some disadvantages when the quantity and optimization interval of inverter controller parameters increase, such as easily falling into local optimum, and optimization convergence problems. An improved PSO algorithm with multi-swarm and multi-velocity (MMPSO) is proposed in this paper. It is made of three kinds of different particle groups and three kinds of velocity update methods. Comparing with the traditional PSO algorithm, the combination of multi-swarm and multi-velocity of the MMPSO can effectively improve optimization ability of traditional PSO algorithm. Various off-line optimization models of single phase inverter systems are established to compare MMPSO and other PSO algorithms. The experimental results validate the validity and correctness of MMPSO.