A combined distribution and assignment model for continuous facility location problem

Abstract. Consider a general, heterogeneous geographical space with a set of competitive facilities, where the customers' demand locations from each of the facilities are continuously dispersed over the area. The total demand generated from a particular location in the space is fixed, but the demands from this location to the set of competitive facilities are subject to a distribution function with respect to the relative transportation costs to these facilities. Furthermore, we take into account congested transportation cost in characterizing customer choices. Congestion effect is explicitly built into our model by using a flow-dependent and location-dependent transportation cost function. The routing behavior of customers over the space and the user equilibrium choices of facilities are modeled by constructing a spatial user equilibrium flow pattern. The problem is formulated as a combined distribution and assignment model. An iterative algorithm between the distribution function for the choice of facilities and a mixed finite element method for route choices is proposed to solve the resulting continuous facility location problem. A numerical example is given to demonstrate the effectiveness of the proposed methodology.