The Improvemence And Application Of Imperialist Competitive Algorithm

Imperialist Competitive Algorithm(ICA)is proposed by Atashpaz-Gargari and Lucas,belongs to the swarm intelligence algorithm.The algorithm is derived from the simulation of the evolution of countries in human society.Individuals in the population are regarded as "countries".In the population,according to the fitness,several countries are selected as colonial countries,while other countries are regarded as colonies.By ingeniously setting up the aggression and competition mechanism,the optimization problem can be solved.As a emerging swarm intelligence algorithm,ICA has attracted the attention of researchers in this field since it was proposed,and has carried out research on performance optimization,exploring new applications and so on.The algorithm has been applied in many fields,such as engineering,logistics,electricity,medicine and so on.Relevant studies show that ICA still has some problems,such as rapid decline of population diversity,easy premature convergence and low resolution accuracy.In order to overcome these weaknesses and improve the performance of the algorithm,this paper makes a thorough study of the algorithm and proposes three improved algorithms.They are: imperialist competitive algorithm with double assimilation mechanisms(DAICA),imperialist competitive algorithm with adaptive competition(AICA),imperialist competitive algorithm using Opposition-Based-Learning(OBLICA).The main research work of this paper is as follows:(1)An imperialist competitive algorithm with double assimilation mechanisms is proposed.ICA algorithm adopts greedy update strategy,so the population tends to gather around the optimal individuals in the later stage,and the state similarity between colonial countries and colonies is high,that is,the diversity of the population decreases,resulting in premature convergence of the algorithm.In order to keep the diversity of the population better in the later stage,the idea of teaching and learning optimization algorithm is used for reference,and the mutual learning strategy between colonial countries is introduced,so that the algorithm has a certain diffusion ability in the later stage and enlarges the search scope of the population.The simulation experiments on Benchmark function show that the imperialist competitive algorithm with double assimilation mechanisms has a significant improvement in search accuracy and stability compared with the standard ICA.(2)An imperialist competitive algorithm with adaptive competition is proposed.In order to improve the accuracy of the algorithm,an adaptive competition coefficient is set up in ICA.The parameter changes according to the number of iterations,and the early stage ofthe algorithm is a smaller value,so that the diversity of the population can be maintained.With the increase of the number of iterations,the value gradually increases,so as to accelerate the convergence speed of the algorithm.Finally,the experiments show that the imperialist competitive algorithm with adaptive competition has better convergence speed and solution accuracy than the standard ICA.(3)ICA is used to solve job shop scheduling problem.In order to solve the job shop scheduling problem better,the ICA is improved.After the imperial competition process,the colonial reform operation is added,and the Opposition-Based-Learning of the colonies is introduced,which strengthens the ability of the algorithm to explore new solutions.The experimental results show that the imperialist competitive algorithm using Opposition-Based-Learning has excellent performance in stability and accuracy.