Generating alternatives for siting retail and service facilities using genetic algorithms and multiple criteria decision techniques

Abstract

Fundamental step in the structuring of siting problems in general and the siting of retail and service facilities in particular is the generation of alternatives. This should occur at the beginning of the process of facility location. The identification of a manageable number of alternatives aims at providing a preliminary insight into the feasibility of the project in the area of concern. These alternatives then undergo further formal analysis. The purpose of this paper is to present a methodology aimed at generating alternatives for siting retail and service facilities, taking into account a number of criteria that reflect different aspects of the problem. This tool was originally set up as the first step in a decision support system for generating and evaluating alternative sites for facilities. It has been developed as a tool for a geographic information system (GSIS). The search is carried out by applying genetic algorithms (GAs) that are analogous to natural processes for optimization and search procedures. According to the algorithm's terminology a fitness function measures the worth of each candidate alternative, which is codified into a chromosome. The merging of aspects of multicriteria theory and of GAs will be useful for the generation of alternatives in location problems. The aim of this present work is to improve the theoretical principles upon with the fitness function is based, leading to the construction of a robust set of alternatives.