Adaptive region adjustment to improve the balance of convergence and diversity in MOEA/D.

Highlights • The balance of convergence and diversity are analyzed and divided into horizontal imbalance and vertical imbalance. • The one-to-one correspondence between the region and the subproblem is established by region division. • Adaptive Region Adjustment (ARA) strategy is proposed to improve the balance between convergence and diversity. • A comprehensive experiment is designed on UF and MOP to prove the effectiveness of our proposed algorithm. Abstract The multiobjective evolutionary algorithm based on decomposition (MOEA/D), which decomposes a multiobjective optimization problem (MOP) into a number of optimization subproblems and optimizes them in a collaborative manner, becomes more and more popular in the field of evolutionary multiobjective optimization. The mechanism of balance convergence and diversity is very important in MOEA/D. In the process of optimization, the chosen solutions must be distinctive and as close as possible to the Pareto front. In this paper, we first explore the relation between subproblems and solutions. Then we propose the adaptive region adjustment strategy to balance the convergence and diversity based on the objective region partition concept. Finally, this strategy is embedded in the MOEA/D framework and then a simple but efficient algorithm is proposed. To demonstrate the effectiveness of the proposed algorithm, comprehensive experiments have been designed. The simulation results show the effectiveness of our proposed algorithms. [ABSTRACT FROM AUTHOR]