| Summary: | 碩士 === 國立中正大學 === 企業管理所 === 95 === Under the global pressure, strengthen the supply chain management became the key issue for individual firms to increase its core competitions. However, from the past efforts of supply chain management, it focused on single base objective, single- echelon supply chain network, single-period decision or even creates a planning horizon under demand certainty. But what was mentioned is no longer useful to the conditions firms are confronting nowadays. Therefore in this study, we came up with multi-objectives, multi-echelons, multi-periods and other conditions to make the planning horizon more effective and efficient to confront match the conditions of fluctuated demand.
This research brought out two models; the first model is formed under the condition without the criterion of risk-pooling while the second model is formed under the concept of risk-pooling. After analyzing the two models, we generate simulation data under the assumption of normal distributed demand to test the two proposed models and bring the calculated results to a pair-wise t test to testify their significances. The analyzed results included the total cost and its other relevant costs such as the purchasing cost, the production cost, the setup cost, the transportation cost, the inventory cost, and the penalty cost for products over and under sent to customers. And the decisions to the supply chain network modeling include the order quantity of raw material, the production amount of each product and the distribution amount of products to each downstream.
After the test, we found out that the model of no risk-pooling tends to be higher than the models of risk-pooling for all scenarios of risk-pooling. To specify the result, the model under risk-pooling, the higher the demand fulfillment is the higher the total cost is. And the majority cost of the total cost did not lay in the penalty costs for products over or under sent to the customers; majority cost are formed according to the objective function which included the purchasing cost, the transportation cost, the production cost and the setup cost. The results above are based on a normal distributed demand assumption.
|
|---|
