A Cultural Algorithm Applied in a Bi-Objective Uncapacitated Facility Location Problem

Abstract

Cultural Algorithms (CAs) are one of the metaheuristics which can be adapted in order to work in multi-objectives optimization environments. On the other hand, Bi-Objective Uncapacitated Facility Location Problem (BOUFLP) and particularly Uncapacitated Facility Location Problem (UFLP) are well know problems in literature. However, only few articles have applied evolutionary multi-objective (EMO) algorithms to these problems and articles presenting CAs applied to the BOUFLP have not been found. In this article we presents a Bi-Objective Cultural Algorithm (BOCA) which was applied to the Bi-Objective Uncapacitated Facility Location Problem (BOUFLP) and it obtain an important improvement in comparison with other wellknow EMO algorithms such as PAES and NSGA-II. The considered criteria were cost minimization and coverage maximization. The different solutions obtained with the CA were compared using an hypervolume S metric.