动态环境下一种改进的小生境粒子群算法  

Improved niche Particle Swarm Optimizer in dynamic environment

作　　者：李孝源[1] 李枚毅[1] 宋凌[1] 

机构地区：[1]湘潭大学信息工程学院,湖南湘潭411105

出　　处：《计算机工程与应用》2008年第9期51-54,共4页Computer Engineering and Applications

基　　金：湖南省自然科学基金(the Natural Science Foundation of Hunan Province of China under Grant No.06JJ5106)

摘　　要：为了提高在动态环境下追踪变化的极点的可靠性和精确性的能力,避免算法收敛于一个最优解,提出了一种改进的小生境微粒群算法。使用DF1(Dynamic Function1)生成的复杂动态环境对这种算法进行了验证,并与经典的APSO(Adaptive Particle Swarm Optimizer)算法进行了对比,实验结果表明了该算法的有效性。The purpose of this paper is to present a modified Panicle Swarm Optimization（PSO ） algorithm applied to the complex dynamic environment.The method presented is defined as Improve Niche Panicle Swarm Optimizer（ IN-PSO ）.It can improve the reliability and accuracy while tracking dynamic pole in dynamic environment and avoid converge to a optimality.The environment used in these experiments is generated by Dynamic Function #1（DF1）.The results of the experiments elucidate that INPSO is more adaptive than Adaptive Panicle Swarm Optimizer（ APSO ）.

关 键 词：动态环境 小生境 微粒群算法 DFL 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]