基于模糊C均值聚类的锦标赛选择机制与多目标优化研究

摘要

本文提出了一种用于多目标优化的进化算法——基于模糊C均值聚类的进化算法(A Fuzzy C-Means Clustering Based Evolutionary Algorithm,FCEA).在算法的迭代过程中,先利用模糊C均值聚类算法寻找种群的分布结构,通过对每一代种群进行模糊划分,获得每个个体隶属于每一类的隶属度,然后本文设计了一种基于隶属度的锦标赛选择算子,用于从整个种群中选择相似个体进行重组,引导算法进行搜索.实验结果表明,基于隶属度的锦标赛选择算子的应用能够提升算法的性能,与MOEA/D-DE、NSGAII、SPEA2、SMS-EMOA等先进的优化算法进行比较的结果表明,FCEA在求解具有复杂Pareto前沿的多目标优化问题(GLT系列)时具有一定的竞争力.%A fuzzy C-means clustering based evolutionary algorithm called FCEA was proposed to optimize multiobjective optimization problems.In the process of iteration of this algorithm,a fuzzy C-means clustering method is firstly employed to implement a fuzzy partition of the population so as to discover the population distribution structure and to obtain a membership matrix of the population at each generation.According to the distribution structure,a membership based tournament selection strategy (MBTS) is designed to select neighboring solutions from the population for recombination and to guide search.The experiments present that MBTS significantly contributes to the performance of FCEA.Comparison experiments show that the proposed FCEA outperforms MOEA/D-DE,NSGAII,SPEA2 and SMS-EMOA on solving GLT test suite with complicated Pareto Front (PF) shapes.