Abstract: This paper intends to develop an improved particle swarm optimization (PSO) algorithm. The proposed method will introduce "Fin}Tuning" into the PSO algorithm which can promote the ability of local search to modify the defects of high similarity of individual particles on the late period of search following PSO algorithm. At last the performance of the improved PSO and PSO will be compared by optimizing five massively multimodal functions with varying complexities. The results show that the performance of the improved PSO is better than PSO on search success rate, average convergence time and average convergence generations.

Key words: Particle swarm optimization, Fin}tuning mechanism, Multimodal functions