In this paper, we consider a form of multi-item inventory management known as the Joint Replenishment Problem (JRP). In our model, the target warehouse sells multiple items to meet the retailer's demand and replenishes these items from the supplier. At each replenishment, the warehouse decides which items should be ordered and how much should be ordered. Carriers with a finite capacity transport items from the supplier to the warehouse with a fixed lead-time. The fixed ordering cost of each carrier is charged according to the number of carriers used, not the order volume. Because of the stepwise ordering cost, full-load replenishments reduce the cost, even if an unscheduled item is ordered. In addition, if a relation exists among the items demanded, the can-order policy is more suitable. Under the can-order policy for JRP, some items are re-ordered when the inventory volume is below the re-order level and any items with an inventory volume below the can-order level can be included in the order. Several studies have considered how to set the parameters of the can-order policy. But, correlated demands have not received sufficient attention in the literature because it is difficult to find the optimal parameters. In this paper, we propose to find the optimal parameters of the can-order policy, the can-order level and the order-up-to level for each correlated demand item by applying a genetic algorithm. The main objectives in our model are minimizing item storage, stock-outs and carrier fees. In numerical experiments, we simulated inventory movement and ordering items while considering the capacity of the carriers. The experimental results of our proposed model are compared with the results of other models.
抄録全体を表示