Research And Improvement Of Artificial Bee Colony Algorithm

With the continuous expansion of applications and requirements,such as non-convex,non-linear,high-dimensional,multi-variable and multi-objective and other complex optimization problems arise,for the optimization of such problems,the traditional optimization algorithm is no longer applicable.The group intelligent optimization algorithm compensates the deficiency of the traditional algorithm with its unique optimization mechanism and is widely used to solve the complicated optimization problem in various fields.Common algorithms are genetic algorithms,particle swarm optimization,ant colony algorithm,artificial fish swarm algorithm.Karaboga proposed an Artificial Bee Colony Algorithm in 2005.The algorithm has the characteristics of simple and easy implementation,low control parameters,strong robustness and strong global search ability.Therefore,it has received the attention of a large number of scholars,and in the real life of production problems are widely used.Firstly,this paper briefly introduces the development of optimization problem and optimization method,Artificial Bee Colony Algorithm is a new group of intelligent optimization algorithms,which has been paid attention to and applied by a large number of scholars since it was put forward.However,the algorithm is still in the initial stage,and there is still "precocious" convergence,weak local search ability and late slow convergence rate and other shortcomings.Based on the analysis of the shortcomings of the algorithm,an improved artificial bee colony algorithm is proposed.The main work is as follows:(1)The theoretical basis,basic principle and implementation steps of the standard Artificial Bee Colony Algorithm are analyzed in depth,and the advantages and disadvantages of the algorithm are summarized based on the above analysis;(2)In view of the shortcomings of the algorithm,this paper mainly from two aspects of the standard artificial bee colony algorithm to improve,first,the use of reverse learning initialization method to increase the diversity of solutions,the second is introduced by differential evolution algorithm inspired search equation,in order to improve the development ability of the algorithm;(3)Through simulation experiments,the improved algorithm has better performance and better optimization ability.