摘要
针对基本粒子群优化算法的早熟收敛问题,提出了一种基于逻辑自映射的变尺度混沌粒子群优化算法。该算法在粒子群优化算法每次寻优结束时,采用逻辑自映射函数产生混沌序列,在已搜索到的精英粒子附近尝试搜索更优解并动态收缩搜索范围,在防止算法过早陷入局部最优的同时提高了算法搜索的精度。仿真结果表明,新算法在寻优成功率和平均最优值方面有很大提高,在求解包括欺骗性函数和高维函数在内的多种函数优化问题方面具有良好的效果。
Aiming at the standard particle swarm optimization（PSO） existing shortcomings of premature and low convergence,this paper proposed a novel altorithm which used the method of mutative scale chaos optimization based on self logical mapping function（SLM-PSO）.In SLM-PSO,computed a series of chaotic variables according to self logical mapping function at the end of each iteration,then the SLM-PSO algorithm attempted to search the better solutions around current best solutions by chaos optimization and shrink search field dynamically.The simulation results for benchmark functions suggest that the new proposed algorithm has better probability of finding the global optima and mean best values,especially for complex multimodal functions.
出处
《计算机应用研究》
CSCD
北大核心
2011年第8期2825-2827,共3页
Application Research of Computers
基金
国家教育部人文社会科学规划基金资助项目(10YJA630187)
高校博士点专项科研基金资助项目(20093120110008)
关键词
逻辑自映射
混沌
变尺度
粒子群算法
函数优化
self logical mapping function
chaos
mutative scale
particle swarm optimization
function optimization
