具有小世界网络搜索能力的智能微粒群算法  

Intelligent Particle Swarm Algorithm with Small-World Network Neighborhood Topology

作　　者：陆虎[1] 李永忠[1] 

机构地区：[1]江苏科技大学计算机科学与工程学院,江苏镇江212003

出　　处：《微电子学与计算机》2009年第12期160-163,共4页Microelectronics & Computer

基　　金：江苏省教育厅高校基金指导计划项目(08KJD520007)

摘　　要：微粒群算法中微粒的邻域搜索范围具有典型的小世界网络特征.文中借鉴小世界网络的生成过程,提出了具有小世界网络搜索能力的智能微粒群算法,将全局寻优位置gbest的更新过程看作为随社会网络结构优化过程的传递,实现了微粒邻域结构的动态变化.从实验结果看出,在不同的测试函数上改变微粒的邻域结构能得到不同的结果.Neighborhood topologies of particle in particle swarm optimization similar to small- world network. Inspired by the structural process of small - world network, a new intelligent particle swarm algorithm with small - world network neighborhood topology is proposed. The updating of globe best space was considered as a process related to optimization of the structure of the social network, which realizes dynamic changes of neighborhood during generations. Experiments of testing functions show neighborhood topology significantly affects the performance of particle swarm optimization.

关 键 词：微粒群算法 小世界网络 邻域搜索 拓扑结构 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]