Convergence Analysis of Cuckoo Search by Creating Markov Chain  

作　　者：ZHOU Hui CHENG Ya-qiao  LI Dan-mei XU Chen 周晖;程亚乔;李丹美;徐晨(College of Electronic Information, Nantong University, Nantong 226019, China College oflnformation Science and Technology, Donghua University, Shanghai 201620, China)

机构地区：[1]College of Electronic Information,Nantong University,Nantong 226019,China [2]College of lnformation Science and Technology,Donghua University,Shanghai 201620,China

出　　处：《Journal of Donghua University(English Edition)》2016年第6期973-977,共5页东华大学学报（英文版）

基　　金：National Natural Science Foundation of China(No.61174065)

摘　　要：Cuckoo search(CS) has been used successfully for solving global optimization problems.From a theoretical point of view,the convergence of the CS is an important issue.In this paper,convergence analysis of CS was studied.The transition probability characteristics of the population to construct a Markov chain were analyzed.The homogeneity of the Markov chain was derived based on stochastic process theory.Then it was proved to be an absorbing state Markov chain.Consequently,the global convergence of CS was deduced based on conditions of convergence sequence and total probability formula,and the expected convergence time was given.Finally,a series of experiments were conducted.Experimental results were analyzed and it is observed that CS seems to perform better than PSO.Cuckoo search（CS） has been used successfully for solving global optimization problems. From a theoretical point of view, the convergence of the CS is an important issue. In this paper, convergence analysis of CS was studied. The transition probability characteristics of the population to construct a Markov chain were analyzed. The homogeneity of the Markov chain was derived based on stochastic process theory. Then it was proved to be an absorbing state Markov chain. Consequently, the global convergence of CS was deduced based on conditions of convergence sequence and total probability formula, and the expected convergence time was given. Finally, a series of experiments were conducted. Experimental results were analyzed and it is observed that CS seems to perform better than PSO.

关 键 词：cuckoo search(CS) global convergence Markov chain expected convergence time 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]