The Multi-Supplier Multi-Depot Vehicle Routing Problem

Abstract

How to decide the optimal routes through which corresponding cargoes from suppliers to logistics centers as well as from logistics centers to customers can be smoothly transported is constantly a big issue in logistics chains. This paper is concerned with constructing the optimal distribution chain (or network) comprised of supply points, logistics centers and customer sites, which is called the Multi-Supplier Multi-Depot Vehicle Routing Problem (MSDVRP). The MSDVRP locates logistics centers, assign vehicles to centers where they station and construct sequences of delivery routes for the vehicles so as to minimize the total relevant cost. The Simulated Annealing (SA) based method to obtain a good approximate solution is developed. The performance of the proposed solving method is analyzed by comparing solutions for small sample problems by SA with exact solutions by a commercial optimal solver, LINGO. Also, the potential adaptability of the SA method is evaluated by some case studies with relatively large scale of problem settings. As a result, it was found that the SA is practically useful for relatively large problems.