An efficient approach to discrete Multiple Different Facility Location Problem

The multiple facility location problem (MFLP) is to locate certain facilities so as to serve optimally a given set of customers, whose locations and requirements are known. When the facility locations have to be selected from a given set of candidate locations, the corresponding location problem becomes a discrete multiple facility location problem (DMFLP). In this study we consider a special case of DMFLP where multiple facilities that are of different type are to be placed (location decision) and assign customers to these facilities (allocation or assignment). The case will be discussed based on with and without interactions among new facilities. We then propose new heuristic solution methods and branch-and-bound algorithms. Computational results on randomly generated data in comparison with optimal solutions indicate that the new methods are both accurate and efficient.