Multi-objective variation differential evolutionary algorithm based on fuzzy adaptive sorting

Abstract

In order to improve the convergence and diversity of multi-objective differential evolutionary algorithm in solving problems, a fuzzy adaptive sorting variation multi-objective differential evolution algorithm is proposed. First of all, using an adaptive fuzzy system by adjusting the parameters of the sorting variation, the balance of local search ability and global exploring ability of the algorithm, at the same time of accelerate the algorithm convergence speed, reduce the possibility of a fall in local optimum; Secondly, using the homogeneous population initialization method, based on the distribution of the algorithm was beginning to get a uniform initial population, improving the stability and diversity; Finally, add a temporary population to store is discarded by individuals, the optimized choice finally, for each generation to improve the population diversity in the process of evolution. Matlab was used to conduct simulation experiments and compared the proposed algorithm with four other multi-target evolutionary algorithms. The experimental results show that the proposed algorithm is superior to several other contrasting algorithms in convergence and diversity, and can effectively approach the frontier of real Pareto. At the same time, the experiment also verifies the validity of fuzzy adaptive sort variation strategy in the proposed algorithm.