Integration Exploration Of Differential Evolution And Improvement And Application Of Its Mutation Strategy

Differential evolution algorithm is a heuristic evolutionary algorithm that simulates the behavior of biological population evolution.It has the advantages of simple process,less control parameters and easy implem entation.Since been put forward,it has been widely concerned and applied to various aspe cts.However,it also has the problem s of poor local search ability and easy premature convergence,which need to be improved.In order to im prove it,this pape r mainly studies the follo wing aspects: With the advantage of some excellent strategies,such as reverse learning and Cauchy m utation,we try to com bine differential evolution algorithm with experim ental research and verify it.T he characteristics of all kinds of mutation st rategies are explored,the mutation operators are im proved and i nnovating,and the im proved differential evolution algorithm is simulated and applied to solve the r esource-constrained project scheduling problems.The specific work includes:The first,an im proved differential evolution algorithm based on Cauchy m utation and opposition-based learning is proposed.Using the Cauchy m utation as a single mutation strategy,and combining with the opposition-based learning strategy,it makes the performance of both the pros and cons of complementary.The algorithm,selects the generation jump strategy in a certain jum p probability random ly and integrates the global exploration ability of the opposit ion-based learning strategy,the lo cal exploitation ability and disturbance function of the Cauchy mutation strategy,let it guide evolution better.Secondly,a differential evolution algorithm based on adaptive mutation operator is propose d.The definition of indivi dual vector pa rticle and dimensional layer is presented.Based on the different dimension's selection strategy for weighted dimensional layer,the weighted di fferent dimensional learning is introduced into differential evolution algorithm for the first time,which can effectively improve the diversity of the population.According to the aggregation degr ee of individual population,and an adaptive m utation operator based on aggregation degree of individual population is proposed.The opera tor can adaptively adjust the variation weight of DE/best/1 mutation operator and the different dimensional learning mutation operator according to the aggregation degree of individual population currently,improves the conver gence speed,of the algor ithm.Finally,the dif ferential evolution algorithm based on adaptive mutation operator is applied to solve the resource-constrained project scheduling problem.The adaptive mutation operator based on population aggregation has strong population diversity,and is not easy to fall into local optimum and has fast conver gence speed.It can find the optim al project scheduling quickly and efficiently.