摘要
提出了一种新的粒子群优化算法(PSO)———带边界变异的PSO,其原理是:在迭代的过程中,当粒子的位置超出可行域时,带边界变异的PSO让粒子的位置重新均匀分布在边界附近;当粒子的速度超出可行域时,则使其均匀分布到整个可行区间,而不是像原始PSO那样在这2种情况下都只是简单地取边界值。从理论和实验2个方面论证了这种引入了边界变异的PSO可以获得更快的寻优速度和更好的解精度,有一定的推广价值。
A new particle swarm optimization （PSO）, PSO with bounded mutation operator, was developed. In contrast to the original PSO, in this algorithm, PSO with bounded mutation operator would make the particles redistributed cquably close to boundary when place of particle exceeds feasible region during iteration. The particles would be redistributed equably in whole feasible region when speed of particle exceeds feasible region. PSO with bounded mutation operator could achieve faster speed of optimization and better precision of optimization, which was proved by the theory and experimental methods.
出处
《武汉理工大学学报》
CAS
CSCD
北大核心
2005年第9期101-103,114,共4页
Journal of Wuhan University of Technology
基金
交通部博士基金(200332581106).
关键词
PSO
边界变异
优化
PSO
bounded mutation
optimizalion
