Ant based heuristics for the capacitated fixed charge location problem

Abstract

This paper presents two different Ant System ( AS) based algorithms for the Capacitated Fixed Charge Location Problem (CFCLP) which is a discrete facility location problem that consists of selecting a subset of facilities that must completely supply a set of customers at a minimum cost. The first algorithm is concerned with extending and improving existing work primarily by introducing a previously unconsidered local search scheme based on pheromone intensity. Whilst, the second method makes a transformation of the derived AS algorithm into the hyper-cube famework in an attempt to improve efficiency and robustness. Computational results for a series of standard benchmark problems are presented and indicate that the proposed methods are capable of deriving optimal solutions for the CFCLP. © 2008 Springer-Verlag Berlin Heidelberg.