压缩搜索空间与速度范围粒子群优化算法  

Particle Swarm Optimization with Contracted Ranges of Both Search Space and Velocity

作　　者：陈炳瑞[1] 冯夏庭[1] 

机构地区：[1]东北大学资源与土木工程学院,辽宁沈阳110004

出　　处：《东北大学学报（自然科学版）》2005年第5期488-491,共4页Journal of Northeastern University(Natural Science)

基　　金：国家重点基础研究发展规划项目(2002CB412708);国家杰出青年科学基金资助项目(50325414)

摘　　要：为了改善粒子群优化(PSO)算法的搜索性能,提出一种改进的粒子群算法CSV PSO算法·该算法在粒子群进化的过程中根据粒子群的最佳适应值动态地压缩粒子群的搜索空间与粒子群飞行速度范围;针对PSO算法可能出现的暂时停滞现象,引入分区重新初始化机制·数值仿真结果表明:随着粒子群进化,适当的压缩粒子群搜索空间与飞行速度范围,有利于加速算法收敛,提高收敛精度;该算法收敛速度更快,精度更高,运行更为稳定·To improve further the performance of PSO(Particle Swarm Optimization), a modified PSO algorithm is proposed and called CSV-PSO algorithm. Based on the best fitness of the particles, the ranges of both search space and velocity of the particles are contracted dynamically with the evolution of particle swarm in CSV-PSO algorithm. To avoid the possible occurence of stagnation phenomenon in the PSO algorithm, the re-initialization mechanism based on different search spaces is introduced in the CSV-PSO. Numerical examples show that it is of advantage to accelerating the algorithm's convergence and improving its calculation accuracy so as to contract appropriately the ranges of both search space and velocity of particles in evolutionary progress and the algorithm is easier for convergence, more accurate for calculation and more stable for running.

关 键 词：粒子群优化 群智能 进化计算 随机优化 自适应 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]