Improving NSGA-II for solving multi objective function optimization problems

It is important for any MOEA (Multi Objective Evolutionary Algorithm) to improve convergence and diversity of solutions of Pareto front, which is obtained at the termination of MOEA. There are many MOEA available in the literature: NSGA-II, SPEA, SPEA2, PESAII and IBEA. This paper aims at improving solutions diversity of Pareto front of a well known multi-objective optimization algorithm, NSGA-II. The standard NSGA-II algorithm uses crowding distance based method for maintaining solutions diversity. The limitation of crowding distance based method is that it selects two nearer solutions from the Pareto front for the mating. The SPEA algorithm uses agglomerative hierarchical average linkage based clustering method for maintaining solutions diversity. The method sometimes may not preserve extreme solutions in Pareto front. In this paper, we propose a new diversity method based on agglomerative hierarchical clustering with extreme solutions preservation. The proposed method is tested on standard test problems of MOEA. It is observed that the proposed method gives good solution diversity on two objectives test problems compared to the existing diversity method of NSGA-II.