Cuckoo Search Algorithm Based On Contraction Factor Strategy

Through the rational use of resources in the system,the optimization problem can achieve the best results with the lowest cost and the best solution.Because of the wide applicability of the optimization problem,it has been applied in many fields.Swarm intelligence optimization algorithm is a heuristic probabilistic search algorithm based on bionics.This algorithm is a new evolutionary computation method,which has the characteristics of indirect communication,robustness,self-organization and easy realization.Swarm intelligence optimization algorithm solves the optimization problem by simulating the animal or insect.Cuckoo search algorithm is a new meta heuristic swarm intelligence optimization algorithm,which solves a series of continuous optimization problems by simulating the search behavior of cuckoo eggs in the nest of the host.Cuckoo search algorithm has the characteristics of few parameters,easy realization,good robustness,Although the algorithm itself has some shortcomings,such as population activity is not high,convergence rate is limited,solution accuracy is not high and so on.This article researches the swarm intelligence optimization algorithm,focuses on the analysis of the classical cuckoo search algorithm,introduces the contraction factor strategy and proposes an improved cuckoo serach algorithm(CFCS).In the initialization phase,an improved Tent chaotic inverse learning strategy is introduced.Improved Tent chaotic sequences can Initialize each dimension's nest location,which can make the nest location is more evenly distributed in each dimension of the space.The improved reverse learning strategies can be generated for each bird's nest of position reverse solution.According to the fitness formula,we can get a group of optimum location of the nest.Therefore,the improved Tent chaotic backward learning strategy can improve the activity and convergence speed of the population.In the update phase,the contraction factor strategy is introduced.In the early stage of evolution,the contraction factor strategy makes the algorithm to search by the optimal step size,which improves the convergence speed of the algorithm,In the later stage of evolution,the contraction factor strategy makes the algorithm fast to the optimal solution,which is helpful to improve the convergence speed of the algorithm.Finally,an selection strategy based on fitness ranking is introduced.The fitness function transforms the objective function value,which differential adjusts the discovery probability of nest position by the fitness value.This strategy is conducive to update the nest position that has optimal objective function value,enhance the accuracy of convergence algorithm.Occasionally update the value of the objective function of poor nest location,is conducive to the algorithm to jump out of local optimum.In this article,five kinds of single peak test functions and four kinds of multimodal test functions are used to experiment CFCS and the second generation of CS.The experimental results show that the CFCS algorithm is superior to the second generation of CS in performance,which improves the convergence speed and global optimization ability of the algorithm,and to some extent,avoids the situation of falling into local optimum.The main research direction of this article is the improvement of cuckoo search algorithm.Evenly initializing the nest position,adaptively adjusting step,dynamically adjusting discovery probability of nest position,are all the improved method influencing the performance of the algorithm.In the follow-up work,I will study these key factors in order to better change the performance of CFCS,and expect to apply the CFCS algorithm in the practical field.