摘要
混沌搜索能够有效跳出局部极小,然而其细搜索能力不足;模式搜索具有很强的细搜索能力,但是其搜索结果的好坏在很大程度上依赖于初始点的选择。为了提高基于混沌搜索的优化方法的搜索精度,基于混沌搜索和模式搜索,本文提出了一种混合混沌模式搜索方法。该方法在混沌搜索的基础上再进行模式搜索得到最终的搜索结果。混沌搜索结果的精度不需要很高,却可以为模式搜索提供有效的初始点,避免搜索陷入局部极小,只需要简单搜索即可得到理想的最优解。仿真结果表明混合混沌模式搜索方法简单、高效。
Chaotic search can effectively jump out of local minima but it has poor fine search ability. Pattern search method has excellent local search ability; however its search result mostly depends on the initial point. To enhance the precision of optimization results, a hybrid chaotic pattern search method （HCPSM） is presented in this paper based on the chaotic search and pattern search method. The final result is found by pattern search based on the result of chaotic search. The result of chaotic search is not required to be high precise, but it can provide an effective initial point for pattern search method to avoid the local minima. The ideal optimum can be found by simple pattern search based on good initial point. Simulation results show that the HCPSM is simple and high effective.
出处
《华东理工大学学报（自然科学版）》
CAS
CSCD
北大核心
2008年第1期126-130,共5页
Journal of East China University of Science and Technology
关键词
最优化
混沌搜索
模式搜索法
optimization
chaotic search
pattern search method
