A MODEL AND CONSTRUCTIVE HEURISTIC APPROACH FOR MULTI-PRODUCT MULTI-PERIOD INVENTORY ROUTING PROBLEM WITH BACKLOGGING AND TIME WINDOW

Today, companies need to integrate all production processes from raw material to nal consumers. Supply chain management suggests opportunities to achieve integration and management within the companies and between them. Since interests of the loops in the supply chain do not necessarily follow the same function, the Vendor-Managed Inventory (VMI) is an approach that seeks to make interaction and coordination between dierent loops in supply chain in the area of inventory and demand management. This paper considers a part of the supply chain that involves a transportation between supplier, customer, and inventory customer management simultaneously by VMI approach to deal with the solution of some kinds of inventory routing problems. The proposed mixed integer linear programming model, in terms of multiple product customer demands, aims to minimize the total costs of transportation, inventory storage, lack of demand, and tardy demands. Due to the complexity of the problem which puts it among NPHard problems, a constructive heuristic algorithm was proposed to solve the model. Two scenarios, each of which consists of 20 samples, were designed to evaluate the performance of the proposed algorithm. Dierent scenarios were created to evaluate the exibility of the objective function in dealing with dierent conditions. Two sets of problems (of small-medium and large-sizes) were presented to evaluate the proposed algorithm. In small- and medium-sized problems, the results of the proposed algorithm compared with those of linear programming model are solved by Cplex solver in GAMS software. To evaluate the performance of the proposed algorithm in producing high quality solutions at small and medium sizes, two methods were proposed. At rst, results of the proposed algorithm are compared with the upper and lower bounds produced by Cplex solver. Second, three performance parameters were de ned and the solutions were evaluated by them. To evaluate the proposed algorithm in large-sized scales, a benchmark Genetic Algorithm was used. Numerical results show the performance of the proposed algorithm.