Hypervolume-Based Multi-Objective Path Relinking Algorithm

Abstract

Abstract. This paper presents a hypervolume-based multi-objective path relinking algorithm for approximating the Pareto optimal set of multi-objective combinatorial optimization problems. We focus on integrating path relinking techniques within a multi-objective local search as an initialization function. Then, we carry out a range of experiments on bi-objective flow shop problem and bi-objective quadratic assignment problem. Experimental results and a statistical comparison are reported in the paper. In comparison with the other algorithms, one version of our proposed algorithm is very competitive. Some directions for future research are highlighted.