计算机应用研究2018，Vol.35Issue(3)：661-665,5.DOI:10.3969/j.issn.1001-3695.2018.03.005
基于多尺度分数阶多重记忆与学习的粒子群算法
Particle swarm optimization algorithm based on multi-scale fractional order multiple memory and learning
摘要
Abstract
In order to improve the problem that the particle swarm optimization algorithm is easy to reduce the population diversity,leading to premature convergence,and fall into the local optimum especially when the particle swarm optimization algorithm(PSO) searches the multi-peak problem in a high dimensional space,this paper developed an improved algorithm which called multi-scale fractional multiple memory and learning PSO (MML-PSO).It introduced the fractional calculus into the velocity and position updating formula,through remembering particle's historic velocity,position trajectory and individual optimal trajectory,population optimal trajectory,it used the long-term memory characteristics of fractional calculus to make full use of the historical information in the process of optimization and improved the convergence speed and convergence accuracy of the algorithm.At the same time,in order to solve some special cases in the process of population evolution,it put forward multi-scale fractional order and correcting trajectories learning strategy to protect the diversity of population and reduce the possibility of falling into local optimum.By compareing MML-PSO with other modified PSO algorithms,the results show that the proposed algorithm has fast convergence speed and high convergence precision.关键词
粒子群优化算法/多尺度分数阶/多重记忆/学习策略Key words
particle swarm optimization(PSO)/multi-scale fractional order/multiple memory/learning strategy分类
信息技术与安全科学引用本文复制引用
董立军,蒲亦非,周激流..基于多尺度分数阶多重记忆与学习的粒子群算法[J].计算机应用研究,2018,35(3):661-665,5.基金项目
国家自然科学基金资助项目(61571312) （61571312）
