Selecting the best model of particle swarm optimization based on the previous performance

Particle swarm optimization (PSO) has been proven to be a simple yet effective algorithm for searching the optimal solutions of objective functions. The main advantage of PSO is its simplicity, but it easily gets stuck in local optima. In order to remain the original merit and raise its performance, a novel idea is proposed in this paper, which selects the best model of PSO based on the previous performance through a scheme of PSO with a switch of multiple models. Experimental results show that the PSO through the scheme outperforms any with its individual setting alone. In the future, PSO algorithms with a switch of multiple models will be a promising research field. In addition, the idea can be easily extended to select the best from multiple optimization methods.