Solution approaches for the integrated production and outbound distribution scheduling problem with multiple plants and perishable items

Abstract

In this manuscript, we consider the integrated production and outbound distribution system in which there are several geographically dispersed facilities. These facilities are responsible for producing and delivering time-sensitive customer demands. We propose a mixed integer programming (MIP) formulation, MIP with an initial solution using Clarke and Wright (MIPCW) and a metaheuristic algorithm (Variable Neighborhood Search-VNS) for the problem whose objective is to minimize the maximum delivery time whereby satisfying all customer demands. We conduct a two-stage experimental study to determine the effectiveness of the proposed metaheuristic algorithm. We compare the VNS algorithm with the best-known solutions for the single plant version of the problem, in the first stage. These results indicate that the VNS algorithm yields promising results. The second stage includes the results for the MIP formulation and the VNS algorithm on a set of newly generated test instances. The average percentage gap between the VNS algorithm and the MIPCW formulation is about 5% which means that the VNS algorithm gives better results in average about 70 s computation time. These results show how effective the suggested VNS algorithm is at solving the test instances.