Constrained multi-objective optimization algorithm with diversity enhanced differential evolution

Abstract

Constrained multi-objective differential evolution (CMODE) is a population-based stochastic search technique for solving constrained multi-objective optimization problems. Although CMODE is a powerful and efficient search algorithm, it frequently suffers from pre-mature convergence, especially when there are numerous local Pareto optimal solutions. In this paper, a diversity enhanced constrained multi-objective differential evolution (DE-CMODE) is proposed to overcome the pre-mature convergence problem. The performance of DE-MODE is evaluated on a set of 8 benchmark problems. As shown in the experimental results, the DE-CMODE performs either better or similar to the classical CMODE.