摘要
研究了求解多峰函数优化问题的粒子群算法.针对粒子群算法易陷入局部最优的缺陷,提出了一种改进方案.该方案将进化过程分为三个阶段,且在每个阶段采用不同的群体规模和惯性权重,并在第一阶段和第三阶段引入了变异操作,以增强算法跳出局部最优的能力.通过对基准函数的测试,结果表明新算法的全局搜索能力有了显著提高,跳出局部最优的能力和其收敛速度明显优于标准粒子群算法.
This paper studies the particle swarm algorithm for solving multimodal function optimization.To overcome the drawback of easily trapping in local optimum,we propose an improved strategy,denoted FPSO.In this strategy,evolution process is divided into three stages,and different group size and inertia are used in each stage.Furthermore,to enhance the ability of jumping out of local optimum,different mutations are introduced into the first and third stage,respectively.The results of simulations for different benchmark functions illustrate that new algorithm improves clearly the global search capability,and the ability of jumping out of local optimum and convergence speed are superior to that of the standard particle swarm optimization.
出处
《陕西科技大学学报(自然科学版)》
2011年第5期140-144,共5页
Journal of Shaanxi University of Science & Technology
基金
国家自然科学基金资助项目(10902062
60671063)
关键词
粒子群算法
阶段进化
变异
早熟收敛
particle swarm optimization
staged evolution
mutation
premature convergence