摘要
针对粒子群优化(PSO)算法的"早熟"现象,给出了基于遗传和粒子群结合的文化演化算法.该算法将PSO/GA纳入文化算法框架,形成PSO的主群体空间和GA的信仰群体空间,两群体空间可以独立并行演化,并在适当的时机实现信仰群体空间对主群体空间的引导,达到改善粒子群优化算法全局搜索能力、提高计算精度的目的.仿真表明,该算法的优化性能和效率优于PSO算法、GA算法和GA-PSO混合算法.
A new cultural algorithm based on the hybrid of GA and PSO algorithm is proposed to solve the ＂Premature＂ problem of global search of PSO algorithm. With both PSO algorithm and GA included in the framework of cultural algorithm, a PSO main population space and GA belief population space are formed, and they can evolve independently in paralled to enable the belief population space to guide the main population space in due time so as to improve the global search ability of PSO algorithm and enhance the computation precision. Simulation results showed that the algorithm proposed is superior to PSO algorithm, GA and GA-PSO hybrid algorithm in optimized performance and efficiency.
出处
《东北大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2009年第11期1542-1545,共4页
Journal of Northeastern University(Natural Science)
基金
国家高技术研究发展计划项目(2006AA060201)
关键词
粒子群优化
文化算法
遗传算法
全局搜索
混合结构
particle swarm optimization （PSO）
cultural algorithm （CA）
genetic algorithm （GA）
global search
hybrid framework
