A Generic Framework for Incorporating Constraint Handling Techniques into Multi-Objective Evolutionary Algorithms

Abstract

A generic framework for incorporating constraint handling techniques (CHTs) into multi-objective evolutionary algorithms (MOEAs) is proposed to resolve the differences between MOEAs from algorithmic and implementation perspective with respect to the incorporation of CHTs. To verify the effectiveness of the proposed framework, the performances of the combined algorithms of five CHTs and four MOEAs on eight constrained multi-objective optimization problems are investigated with the proposed framework. The experimental results show that the outperforming CHT can vary by constrained multi-objective optimization problems, as far as examined in this study.