东北大学学报（自然科学版）

东北大学学报(自然科学版) ›› 0, Vol. ›› Issue (): 0-0.

• 信息科学与工程 •    下一篇

混沌反向学习和声搜索算法

欧阳海滨1,高立群2,邹德旋3,孔祥勇2   

  1. 1. 东北大学
    2. 东北大学信息科学与工程学院
    3. 江苏师范大学 电气工程及自动化学院
  • 收稿日期:2013-03-11 出版日期:2013-09-15 发布日期:2013-04-22
  • 通讯作者: 欧阳海滨
  • 基金资助:
    国家自然科学基金资助项目

Chaos Opposition-based Learning Harmony Search Algorithm

  • Received:2013-03-11 Online:2013-09-15 Published:2013-04-22

RichHTML

0

摘要： 为了改善和声搜索算法易陷入局部最优的不足，本文提出了一种混沌反向学习和声搜索算法（COLHS）。该算法基于算法的聚集和发散思想的考虑，对算法陷入局部最优和停止状态进行初步预判断，并根据预判断的结果融合混沌扰动策略和反向学习，利用Logistic混沌序列的遍历性和反向学习的空间可扩展性，提高算法跳出局部最优的能力。此外，COLHS算法利用和声记忆库的历史信息定义更新因子和进化因子，以自适应地调整参数基音调整概率(PAR)和基音调整步长(bw)，使COLHS算法能够在不同的搜索阶段通过有效地调节参数来平衡算法的聚集和发散。数值计算结果表明，COLHS算法优于HS算法及最近文献报道的8种改进HS算法， 具有良好的全局优化能力。

关键词: 混沌扰动策略, 反向学习, 局部最优, 历史信息, chaos disturbance strategy, opposition-based learning, local optimal, history information

Abstract: Harmony search algorithm(HS) exists a shortcoming is that it is easy to fall into local optimal, For the purpose of improving this shortcoming, chaos opposition-based learning harmony search algorithm(COLHS) is proposed in this paper. Based on the consideration between aggregation and divergence, we preliminary judges this algorithm whether is trapped into local optimal or backwater status, then according to the judge result to integrate chaos disturbance strategy and local opposition-based learning. The ergodic of Logistic Chaos sequence and the space extensibility of opposition-based learning used to make the algorithm escaping local optimal. Beside, COLHS uses the history information of harmony memory to define updating factor and evolution factor. The two factors are applied to dynamically adjust pitch adjustment rate (PAR) and bandwidth (bw), which is aimed at making COLHS able to balance aggregation and divergence by effectively adjust parameter in different search stage. The numerical calculation results demonstrated that the proposed algorithm has better global optimal ability than HS and the other 8 improved HS algorithms were reported in recent literature.