Multi-objective evolutionary algorithm based on similar individuals
Wen Shi-hua · Computer Engineering and Applications Journal · 2008
Diversity maintenance strategy is an important part of studies on multi-objective evolutionary algorithms(MOEAs).A good diversity can give a decision-maker more reasonable and efficient selections.The diversity of pareto optimal solutions include the span and the uniformity of solutions.This paper proposes a multi-objective evolutionary algorithm based on similar individuals(SMOEA).In the process of population maintenance,it deletes individuals with the most similarity.While in the process of evolution operations,the most similar individuals are selected to evolve.Compared with NSGA-II and e-MOEA,the experimental results demonstrate that the new algorithm can converge to the true Pareto front fast and can obtain good convergence at the same time.