人工蜂群算法的收敛性分析.PDFVIP

第28 卷 第10 期 控 制 与 决 策 2013 年 10 月 Vol. 28 No. 10 Control and Decision Oct. 2013 文章编号: 1001-0920 (2013) 10-1554-05 人工蜂群算法的收敛性分析 1,2 1 宁爱平 , 张雪英 (1. 太原理工大学信息工程学院，太原030024 ；2. 太原科技大学电子信息工程学院，太原030024) 摘 要: 利用随机过程理论, 对人工蜂群算法收敛性进行理论分析, 给出人工蜂群算法的一些数学定义和蜜源位置 的一步转移概率, 建立人工蜂群算法的Markov 链模型, 分析此Markov 链的一些性质, 论证了人工蜂群状态序列是有 限齐次Markov 链, 且状态空间是不可约的. 结合随机搜索算法的全局收敛准则, 证明了人工蜂群算法能够满足随机 搜索算法全局收敛的两个假设, 保证算法的全局收敛. 关键词: 人工蜂群算法；Markov 链；转移概率；收敛性 中图分类号: TP301.6 文献标志码: A Convergence analysis of artificial bee colony algorithm NING Ai-ping 1,2 , ZHANG Xue-ying1 (1. College of Information Engineering，Taiyuan University of Technology ，Taiyuan 030024 ，China ；2. College of Electronics and Information Engineering ，Taiyuan University of Science and Technology ，Taiyuan 030024 ，China. Correspondent ：ZHANG Xue-ying ，E-mail ：tyzhangxy@163.com Abstract: The convergence of artificial bee colony algorithm is analyzed theoretically by using the stochastic process theory. Some mathematical definitions of artificial bee colony algorithm and one step transition probability of nectar source position are given and the Markov chain model of the algorithm is established. Some properties of the Markov chain are analyzed, and the conclusions that the artificial bee colony state sequence is a finite homogeneous of Markov chain and the state space of artificial bee colony is irreducible are obtained. It is proved that the artificial bee colony algorithm ensures the global convergence as the algorithm meets two assumptions of the random search algorithm for the global convergence. Key words: artificial bee colony algorithm ；Markov c