Elite Particle Swarm Optimization with mutation

An improved algorithm for Particle Swarm Optimization (PSO) named Elite Particle Swarm Optimization with Mutation (EPSOM) is proposed in this paper. Elite particles and bad particles are distinguished from the swarm after some initial iteration steps. Bad particles are replaced with the same number of elite particles, and a new swarm is generated. To avoid losing diversity of the swarm and to decrease the risk of trapping in local optimum, mutation operation is introduced in evolution process. The results of several simulations for different benchmark functions illustrate that EPSOM algorithm has the ability of local exploitation and global exploration. EPSOM algorithm outperforms the Linearly Decreasing Weight Particle Swarm Optimization (LDW-PSO) and Random Mutation Particle Swarm Optimization (RM-PSO) in respects of calculation accuracy and convergence.