一种高效的改进粒子群优化算法  

High efficient algorithm of modified Particle Swarm Optimization

作　　者：李洪亮[1] 侯朝桢[1] 周绍生[2] 

机构地区：[1]北京理工大学自动控制系,北京100081 [2]曲阜师范大学自动化研究所,山东曲阜273165

出　　处：《计算机工程与应用》2008年第1期14-16,30,共4页Computer Engineering and Applications

基　　金：国家自然科学基金(the National Natural Science Foundation of China under Grant No.60574080)。

摘　　要：提出了一种高效的改进的粒子群优化策略,把整个群体分为几个子群体,进行子群体的专业化社会分工与信息交换,该策略在提高算法局部搜索能力的同时也兼顾了全局搜索能力。测试表明,与现有方法比较,该方法全局寻优的精度与速度有明显提高。A high efficient algorithm of modified particle swarm optimization is presented.Population divided into subpopulations, each subpopulation has its special role and information exchanged between them.This strategy has improved the ability of local search along with the global search ability.Experiments show that this method has been significantly improved on both optimization speed and computational accuracy.

关 键 词：粒子群优化 进化算法 策略 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]