Particle swarm optimization algorithm with multiple social learning structures

This paper proposes a variant of Particle Swarm Optimisation (PSO) algorithm which enhances the social learning structure of the standard PSO by incorporating multiple social best positions. The research in this paper analyses the effects of main parameters on the proposed algorithm's performance by using factorial experiment. To verify the research findings, this paper compares the proposed algorithm's performance to those of several well-known PSO algorithms. Eventually, the comparison results indicate that the proposed algorithm outperforms others.