Research on hybrid PSO algorithm with appended intensification and diversification

The intensification mechanism of basic PSO algorithm was analyzed as well as the diversification mechanism. The problems such as matter correlation influence by particles´ initialization, and algorithm precision were discussed. Then a strategy to improve BPSO was proposed, which could make the search process at the oscillating balance between intensification and diversification. Based on this, a more effective Hybrid Particle Swarm Optimization (HPSO) was proposed, combining the Variable Dimension Greedy Randomized Adaptive Search Procedure (VDGRASP) and the Variable Intension Mutation Procedure (VIMP) with the BPSO. And the experiments of two multi-peak problems as Rastrigin and Schwefel were carried out, which verified the advantage of the new HPSO algorithm.