摘要
以保证全局收敛的随机微粒群算法SPSO为基础,本文提出了一种改进的随机微粒群算法——SM-SPSO。该方法是在SPSO的进化过程中,以单纯形法所产生的最优个体来代替SPSO中停止的微粒,参与下一代的群体进化。这样既可以利用单纯形法的收敛快速性,又可以利用SPSO的全局收敛性。通过对两个多峰的测试函数进行仿真,其结果表明在搜索空间维数相同的情况下,SM-SPSO的收敛率及收敛速度均大大优于SPSO。
Based on the stochastic particle swarm optimization algorithm that guarantees global convergence, an improved stochastic particle swarm optimization algorithm named SM-SPSO is proposed. During the evolution of SPSO, the best particle produced by the simplex method substitutes for the stopping particle, and takes part in the evolution of the next generation. Thus, both the characteristics of speedy convergence of the nonlinear simplex method and the global convergence of the stochastic particle swarm optimization algorithm are used. Through the experiments of two multi-modal test functions, the result of simulation proves that the speed of convergence and the rate of convergence for SM-SPSO are better than SPSO on the same dimension of search space.
出处
《计算机工程与科学》
CSCD
2007年第1期90-93,共4页
Computer Engineering & Science
基金
教育部重点科技项目(204018)
关键词
随机微粒群算法
单纯形法
全局优化
stochastic particle swarm optimization
simplex method
global optimization
