遗传并行粒子群优化算法及其性能分析  

Genetic parallel particle swarm optimization and performance analysis

作　　者：刘昊[1] 李大卫[1] 王莉[2] 

机构地区：[1]辽宁科技大学理学院,辽宁鞍山114051 [2]辽宁科技大学电子与信息工程学院,辽宁鞍山114051

出　　处：《辽宁科技大学学报》2008年第5期495-499,共5页Journal of University of Science and Technology Liaoning

基　　金：辽宁省高校科研计划项目(05L008)

摘　　要：在已有的并行粒子群优化算法的基础上,结合遗传算法,并利用Java语言支持多线程特点,开发出单子群、k子群、任意子群三种遗传并行粒子群优化算法。通过对6个Benchmark测试函数的测试分析,表明这三种算法都具有运行速度快,求解质量高的特点。Based on the existing parallel particle swarm optimization algorithm, combined genetic algorithm, and by using Java language which supports multi-thread, three kinds of parallel particle swarm optimizations of genetic algorithm, which are algorithm of single individual sub-swarm, k sub-swarm and arbitrary sub- swarm, are developed. Through the testing analysis of six benchmark testing functions, it shows that three kinds of algorithm possess the character of quick operating speed and high quality solution.

关 键 词：粒子群优化 并行 遗传算法 Benchmark测试函数 性能分析 

分 类 号：O224[理学—运筹学与控制论]