关键词: 进化算法, 粒子群算法, 全局优化, 慢变函数, 自适应

Abstract: To accelerate searching speed and optimization accuracy of traditional PSO, an improved particle swarm optimization (PSO) algorithm was presented. Regularly varying function and slowly varying function were introduced in the position and velocity update formula. New mechanisms such as explorative operator and exploitative operator are formulated. At the beginning, most elements will be updated by explorative operator in the entire search space sufficiently. Within the iterations, more and more particles will be handled by exploitative operator, which are useful to overcome the deceptions of multiple local optima. It can be seen from the simulation results of the standard benchmark test functions that the proposed algorithm not only accelerates the convergence process, but also improves global optimization ability.