Extension of a Multi-modal Multi-objective Evolutionary Algorithm with Two-objective Problem Transformation to Constrained Problems

Abstract

Multi-modal Multi-objective Problems (MMOPs) commonly arise in real-world problems where distant solutions in the decision space correspond to the same objective values. To obtain all solutions for MMOPs, many Multi-modal Multi-objective Evolutionary Algorithms (MMEAs) have been proposed. In the previous study, we proposed the Multi-modal Multi-objective to Two-objective (MM2T), a decomposition-based MMEA, which has demonstrated superior performance compared to the latest MMEAs. However, the conventional MM2T does not consider constraints. Real-world problems often involve constraints. In recent years, benchmark problems related to Constrained MMOPs (CMMOPs) have been proposed, and research on Constrained MMEAs (CMMEAs) is increasing. In this study, we incorporate four constraint handling techniques into MM2T and evaluate their performance on several CMMOPs benchmark problems. Furthermore, by comparing these variants with the latest CMMEAs, we assess the applicability of the MM2T framework to CMMOPs.