多策略协同进化粒子群优化算法  

Multi-strategy cooperative evolutionary particle swarm optimization

作　　者：张洁[1] 裴芳[1] 

机构地区：[1]湖南机电职业技术学院信息工程系,长沙410151

出　　处：《计算机应用研究》2013年第10期2965-2967,2990,共4页Application Research of Computers

基　　金：湖南省教育厅资助科研项目(12C1065)

摘　　要：为了提高粒子群优化(PSO)算法的优化性能,提出了一种多策略协同进化PSO(MSCPSO)算法。该方法引入了多策略进化模式和多子群协同进化机制,将整个种群划分为多个子群,每个子群中的粒子按照不同的进化策略产生新的粒子。子群周期性地更新共享信息,以加快算法的收敛速度。通过六个基准函数实验,仿真结果表明,新算法在计算精度和收敛速度方面均优于其他七种PSO算法。To improve the optimization performance of particle swarm optimization （ PSO）, this paper proposed a multi-strate- gy cooperative evolutionary PSO （MSCPSO） algorithm. It employed multi-strategy evolutionary pattern and multi-swarm coop- erative evolutionary mechanism. It divided the whole swarm into several sub-swarms, in which each particle generated new particles according to different evolutionary patterns. The sub-swarms updated their sharing information periodically to accele- rate the convergence speed. It conducted experiments on six benchmark functions. Simulation results show that the new ap- proach outperforms seven other PSO algorithms in terms of solution precision and convergence speed.

关 键 词：粒子群优化 多策略 协同进化 全局优化 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]