Proposed a hybrid algorithm for facility location problems

Abstract

The study of facility location problems is directly related to organizational problems of society, such as the location of schools, health centres, etc. In its general form, the problem of p-medians and maximum coverage is NP-Hard, and heuristic methods are used to resolve it. The Differential Evolution algorithms are powerful evolutionary optimization algorithms, originally proposed for problems in continuous spaces. Recently, it has been proposed that adjustments can be made to the mechanism of differential mutation for its application to combinational problems. This paper presents a new hybrid algorithm, using Differential Evolution Algorithms and Tabu Search, to address problems of p-medians and maximum coverage. For the Differential Evolution algorithm certain adaptations are presented, some unpublished, in order to solve the problems in a discrete search space. Computational tests were performed with instances from the litera- 3076 Danielle Durski Figueiredo et al.ture, and the results suggest that the proposed technique is promising and appropriate for the resolution of the problems addressed.