J4 ›› 2014, Vol. 41 ›› Issue (6): 100-105.doi: 10.3969/j.issn.1001-2400.2014.06.017

• 研究论文 • 上一篇    下一篇

一种采用聚类单纯形搜索的引力算法

刘杰1,2;王宇平3   

  1. (1. 西安电子科技大学 数学与统计学院,陕西 西安  710071;
    2. 西安科技大学 理学院,陕西 西安  710054;
    3. 西安电子科技大学 计算机学院,陕西 西安  710071)
  • 收稿日期:2014-02-28 出版日期:2014-12-20 发布日期:2015-01-19
  • 通讯作者: 刘杰
  • 作者简介:刘杰(1977-), 男, 讲师, 西安电子科技大学博士研究生, E-mail: tears191@foxmail.com.
  • 基金资助:

    国家自然科学基金资助项目(61272119, 11301414, 11226173)

Central force optimization algorithm via clustering simplex search

LIU Jie1,2;WANG Yuping3   

  1. (1. School of Mathematics and Statistics, Xidian Univ., Xi'an  710071, China;
    2. College of Science, Xi'an University of Science and Technology, Xi'an  710054, China;
    3. School of Computer Science and Technology, Xidian Univ., Xi'an  710071, China)
  • Received:2014-02-28 Online:2014-12-20 Published:2015-01-19
  • Contact: LIU Jie

摘要:

提出了一种基于聚类和单纯形法的中心引力优化算法,该算法引入了一种聚类单纯形算子,利用聚类方法选择合适的个体以构成单纯形的顶点,周期性地把单纯形算子搜索得到的最优个体迁移到中心引力算法的种群个体中.依靠聚类单纯形算子提高中心引力算法的局部搜索能力和收敛速度,聚类单纯形算子依靠中心引力算法跳出局部最优.实验结果表明,改进后的算法比其他常见优化算法有更好的收敛精度和收敛速度.

关键词: 中心引力优化, 聚类分析, 单纯形, 全局优化

Abstract:

An improved central force optimization (CFO) is proposed based on the clustering and simplex method for global optimization. The clustering simplex (CS) operator is introduced to a new algorithm in the evolution process. Vertices of simplex are selected by clustering methods, and a periodical migrating of the best individual is introduced by the CS operator. CS can get away from local converged points by virtue of CFO, and CFO can improve its local exploiting capability and effectively speed up the convergence under the help of CS. Experimental results show that the proposed hybrid CSCFO algorithm is better than other algorithms in convergent speed and searching precision.

Key words: central force optimization, cluster analysis, simplex, global optimization

中图分类号: 

  • TP301