A New Artificial Bee Colony Algorithm Based On Arithmetic Crossover And Global Search And It's Effectiveness Analysis

Recently,the artificial bee colony(ABC)algorithm,which is a group intelligence algorithm,has been attracting attention as a model of foraging behavior of bees.The ABC algorithm is an optimization algorithm developed for function optimization,and has been verified to be particularly effective for high dimensional problems.However,in the ABC algorithm,there are some problems in the characteristics,such as the necessity of a large number of generations before converging to an excellent solution in order to search for the diversity of the individual.In order to improve this problem,many researches have been made on hybrid evolutionary computation methods to overcome the conventional problems by increasing multiple search methods.In this study,an arithmetic crossover based ABC algorithm(AC-ABC),which incorporates arithmetic crossover,which is one of the crossover methods used in real-time GA for search algorithm of ABC algorithm,is used for the search algorithm of ABC algorithm.This paper proposes a global search type ABC algorithm(GS-ABC),which improves search performance without degrading the nature of local search by combining search processing.In order to verify the effectiveness of the proposed method,we have also simulated the function approximation using various benchmark functions.