A Clustering-Based Approach to the Multi-stage Facility Location Problem

This paper focuses on a special multi-stage facility location problem. The facility is a cluster of equipment that provides service. it is a special facility with high flexibility that can be moved at any time to respond to multi-stage maintenance requirements. This paper aims to find the optimal number of dynamic service facility and the location of each stage, so as to minimize the total cost. In order to solve this problem, two approach are proposed in this paper. The first approach adopt that strategy of front to back, using clustering-based algorithm to get approximate optimal solution; The other is to use a strategy from the back to front, using a dynamic planning algorithm to get the global optimal solution. Numerical experiments are carried out to compare the two approach. The results show that the clustering-based algorithm has higher computational performance and can obtain reliable approximate optimal solution. Finally, we apply the clustering-based approach to real cases to provide a reliable solution for decision makers.