Formulations and branch-and-cut algorithms for multivehicle production and inventory routing problems

摘要

A retail supply chain consists of sequential activities of production, storage, and distribution, each individual process is mostly planned and optimized using predetermined decisions from its preceding activity. In other words each operation is optimized within its framework not taking into account the overall impact on the supply chain as a whole. The real optimization is possible only by treating it as an integrated planning problem. As for inventory is concerned, the total inventory of the supply chain should be optimized instead of optimizing production, whole sale and retail operations separately. This study considers both the IRP and PRP with a discrete and finite time horizon so that it will meet the overall requirements. The number of customers, periods and vehicles are represented by a, b, and c and the instances are represented by ac/bp/cv. Archetti et al. (Ref. 1) developed a branch-and-cut approach for IRP with single vehicle. Archetti et al. (Ref. 2) used the branch-and-cut approach for PRP-ML with uncapacitated production and a single vehicle. This paper considers a single product and a production-distribution network that consists of a production plant and multiple customers that have their own storage area and both may have initial inventory. In each period, each customer need to have sufficient inventory to satisfy its demand. In the case of the PRP, the plant must decide whether to produce the product and the production quantity. For production, fixed setup and unit production costs are involved. The produced quantities can be transported by a limited number of capacitated vehicles to the customers' area, and routing costs are paid. The product can also be stored at the plant or at the customers' area, and unit inventory holding costs are incurred. The cases where the customer replenishment is based on the ML and OU policies are also considered. (22 refs.).