Computer Engineering and Applications ›› 2016, Vol. 52 ›› Issue (19): 25-30.

Previous Articles     Next Articles

Stochastic recursion analysis of particle swarm optimization and its improved algorithm

LUO Jinyan   

  1. Department of Mathematics, Minjiang University, Fuzhou 350108, China
  • Online:2016-10-01 Published:2016-11-18

粒子群算法的递推分析及改进

罗金炎   

  1. 闽江学院 数学系,福州 350108

Abstract: Based on the theory of stochastic recursion process, this paper studies the convergence of the particle swarm optimization algorithm, The parameters of the algorithm and the related conditions are given. Then an improved algorithm with stochastic approximation (RSPSO) is proposed. Simulation results show that RSPSO can effectively avoid the premature convergence and possess more powerful global search capabilities, better performance of optimization.

Key words: Particle Swarm Optimization(PSO), stochastic recursion, premature convergence, stochastic disturbance

摘要: 对带速度项的PSO算法和不具速度项的动态概率PSO算法进行了随机递推分析,给出了保证收敛的算法的参数取值依据以及相关条件,并基于此提出了改进的动态概率PSO算法(RSPSO)。数值实验分析结果表明,改进的PSO算法能有效避免过早收敛,具有较强的全局搜索能力,且优化能力有了进一步提升。

关键词: 粒子群算法, 随机递推, 过早收敛, 随机扰动