融合模式搜索与人口迁移的优化算法  

Optimization Algorithm Combining Pattern Search with Population Migration

作　　者：邓辉[1] 王勇[1] 陈士亮[1] 

机构地区：[1]广西民族大学数学与计算机科学学院,南宁530006

出　　处：《计算机工程》2011年第13期175-177,共3页Computer Engineering

基　　金：广西自然科学基金资助项目(0832084)

摘　　要：针对人口迁移算法存在的收敛速度慢、优化精度低、易陷入局部最优等问题,提出一种融合模式搜索与人口迁移的优化算法。将模式搜索方法作为一种局部优化算子嵌入到人口迁移算法中,对人口迁移算法随机产生的点利用模式搜索方法进行局部优化。通过5个典型函数进行测试,结果表明,该算法收敛速度快、优化精度高,能有效求解全局优化问题。In the light of the deficiencies of Population Migration Algorithm（PMA）,such as slow convergence speed,easily to fall into local optimum,the low solution precision and so on,this paper takes pattern search method as a local optimization operator and embeds it into PMA,that is to say,every point which is randomly generated by PMA is locally optimized by pattern search,and proposes an optimization approach named HJPMA which combines pattern search with PMA.The simulation results show the proposed algorithm has faster convergence speed,higher computational accuracy,and is quite effective for solving global optimization problems

关 键 词：模式搜索 人口迁移算法 优化算法 全局优化 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]