Discretization Based Heuristics for the Capacitated Multi-facility Weber Problem with Convex Polyhedral Barriers

The Capacitated Multi-facility Weber Problem addresses optimally locating I  capacitated facilities in the plane and satisfying demand of J customers so as to minimize the total transportation cost. It assumes that facilities can be located anywhere on the plane and customers are directly connected to them. This study considers the case where there exist convex polyhedral barriers blocking passage and locating facilities inside. Then, the distances between facilities and customers have to be measured by taking into account the polyhedral barriers. The resulting problem is non-convex and difficult to solve. We propose several discretization based heuristic procedures which are especially designed for the problem. The performance of the suggested methods are tested on an extensive set of randomly generated test instances which are derived from standard test instances. Our results imply that the suggested heuristics yield very accurate and efficient solutions for this problem.