一种基于混沌思想和邻域探测机制的PSO算法

A PSO Algorithm Based on Chaos and Self-adaptive Neighborhood Explored

在分析基本PSO算法早熟收敛原因的基础上,提出一种基于混沌思想和自适应邻域探测机制的粒子群优化算法(CANE-PSO).该算法先引入混沌思想对粒子种群进行位置初始化,以提高初始种群的多样性,再引入随机的邻域探测机制,并引入全局极值变异算子,增强了算法的全局搜索能力.通过与其它三个改进算法比较,结果表明CANE-PSO优化效率有较大的提高,较有效地避免了早熟收敛问题.

Based on the analysis of the cause of the premature convergence,a novel particle swarm optimization algorithm based on chaos and self-adaptive neighborhood explored is proposed,which is called CANE-PSO.Chaos is introduced to initialize the particle＇s position to improve the diversity,and the neighborhood detection mechanism and global extreme mutation operator are also introduced to enhance the global search ability.Compared with other three improved algorithms,the CANE-PSO converges faster,and it prevents the premature convergence problem more effectively.