Cooperative Random Learning Particle Swarm Optimization

Particle swarm optimization (PSO) is a recently developed simple and efficient optimization technique and has been applied widely to real life optimization problems. This paper presents an improved version of the original PSO called the cooperative random learning particle swarm optimization (CRPSO), which employs several sub-swarms to seek the space and uses a modified velocity updating equation during the search process. The proposed CRPSO algorithm maintains the diversity of the swarm efficiently and enhances the local search ability simultaneously. The experiment results demonstrate that the CRPSO can improve the performance of the original PSO significantly both on the unimodal and the multimodal function optimization problems.