Theory Research Of Cuckoo Search Algorithm And Its Application

Along with the development of the intelligent algorithm, all kinds of intelligentalgorithm are presented. Metaheuristic is the important component of intelligent algorithm.Such algorithm was inspired by some interesting phenomenons in nature. Cuckoo Search (CS)algorithm is a new Metaheuristic. And it is created by Xin-she Yang and Deb in2009. Thealgorithm has some characteristics, which are faster of searching speed, fewer of usingparameters, better of choosing path, and so on. Now, CS algorithm has been used in neuralnetwork training, multi-objective optimization, software development, and structureoptimization, etc. Because the development of CS has just started, and there is a hugeemptiness of using this method in research area, such as the basis of mathematics theory, thefurther exerting algorithm of derivative algorithms, so CS algorithm has a wide researchprospect. In order to improve research of CS, by Markov chain theory, CS convergence isresearched. And some derivative algorithm of CS is further studied. This paper mainlyresearch contents as follow:(1) The background of the CS algorithm, the basic principle and the characteristics of thealgorithm, and the current development situation will be introduced.(2) Based on the principle of CS algorithm, Markov chain model is established, usingtheory of Markov chain, convergence of CS is proved. And through the simulationexperiment, CS ultimately converge to the global optimal.(3) In order to improve the convergence speed, A Gaussian-disturbance-based cuckoosearch (GCS) algorithm is proposed which adds Gaussian disturbance to search path,through the simulation experiment, the convergence speed of GCS is better than CS.(4) Based on PSO, a hybrid optimization algorithm of CS-PSO is proposed. The searcharea is extended by CS-PSO, and the defect of PSO is easily fall into point of localoptimization that is improved. Through the various of simulation experiments, theresults show that CS-PSO has good ability of optimization.(5) CS is applied to engineering optimization of practical problems;(6) Summarize and prospect.