本研究以動態隨機最佳化來探討封閉式供應鏈。核心元件的主要回收來源為新產品與再製品市場,回收來的核心元件會因為新舊的不同程度而有了品質上的差異,因此具有不確定性。在每一期,企業必須要決定生產多少新產品和再製品以及新舊市場核心元件的回收數量。因此本研究採用連續決策過程(Sequential Decision)來建構整個數學模式,並採用數值迭代法(Value Iteration Algorithm)來進行求解,整個求解的目標是銷售利潤最大化。而整個模型的穩態解可以用來探討庫存水準和各決策之間的關係。最後我們會把動態模型拿來和目前常用的策略進行比較,而動態模型的結果會顯示出在哪些情況下比較好。另外我們會針對不同的參數如何影響決策來進行探討。
This paper presents a stochastic optimization model for a closed-loop supply chain. The core component can be recycled from either new or used market. The quality of recycled core component is uncertain. In each period, the manager has to decide how many new products to produce, how many used products to remanufacture, and how many core components to recycle from new and used market. The objective is to maximize the profit. An efficient value iteration algorithm is developed to solve the problem. The stationary solutions provide relationships between values of decision variables and inventory levels. The performance of dynamic policy is compared with two common practices. The result shows that dynamic policy is desirable in certain conditions. Managerial insights regarding how parameters impact decision variables are discussed.