Particle Swarm Optimization Based on Racing Factor and Mutation Mechanism

Abstract

Abstract A hybrid modified Mutation Mechanism (MM) and Racing factor (RF)-based particle swarm optimization (MM-RF-PSO) are presented in this letter. RF and MM are proposed to determine the approximate location of the global optimal solution. Once the approximate position of the optimal solution is determined, a method to dynamically adjust the parameters of algorithm is proposed, so that the algorithm can approach the optimal solution faster. At the same time, when the algorithm falls into the local optimum, a scheme of jumping out of the local optimum is proposed. Compared with other three improved PSO, the simulation results indicate the effectiveness of MM-RF-PSO.