應用預測存貨模式於專案計畫物料需求之研究－以某研發機構為例

Applying predictive inventory model on the project MRP system: A case study

王東陽(Dong-Yang Wang)

指導教授 ： 張百棧

元智大學/工程學院/工業工程與管理學系/碩士(2004年)

摘要

本研究係以某研發機關之五項關鍵物料之歷史需求為基礎，利用指數平滑法、類神經網路、多元迴歸分析與其合併法探討在需求環境變動下之最適用方法，研究證明合併預測法最適用於個案機構，進而利用此法發展出一套適用個案機構之預測存貨管理模式，預測存貨管理模式為訂購數量與週期皆不固定情況下於接近訂購點(一般模式)時加入預測機制，可減少安全庫存之存貨需求，以取得較佳之訂購時間與數量，經研究證明利用此法可有效降低平均庫存且在顧客服務水準(缺貨率)方面均較個案機構以往使用方法為佳。專案計畫物料隨其專案結束，在缺乏標準化下往往無法應用於其他計畫，物料需求預測之建立更加彌足珍貴，時值個案機構組織再造如火如荼展開，管理者可利用此模式之建立應用於該機構存貨管理。

關鍵字

物料需求預測 ； 存貨管理 ； 倒傳類神經網路 ； 多元迴歸分析 ； 合併法

並列摘要

The research is directed against with the past claim record from the five kinds of the pivot materials by one research and development organization. To take advantage of Exponential Smoothing Method、Artificial neural network、Multiple-regression analysis And combine forecast method, probing into the suitable manner under the variation of the demand condition. The investigation is approved that combine Forecast method is the most suitable for the singular institution. According to combine Forecast method, building the manner of the divinatory inventory control. On the basis of order quantity and periodicity are un-fixed, this manner of the divinatory inventory control is proceeding to make the divination while getting closed the order point (general mode), to reduce the inventory requisition of the secure storage for gaining the better order timing and quantity. This manner is proved to reduce the average secure storage in efficiently, and then the plane of the customer service (the rate in short supply) is better than the original one that an organization used. The materials and supplies for the special case as often as not be applied for the other case once the special case is finished. To build up the divinatory inventory control is required extremely. At present, re-build the organization is starting like a raging fire, the management could utilize and set up this manner for the inventory control.

並列關鍵字

Materiel requirement forecast ； Inventory manager ； Back-Propagation Network ； Multiple-regression analysis ； The combination of forecasts

參考文獻

1.Chan, C. K., B. G. Kingsman, and H. Wang, “The value of combining foreast in inventory management a case study in banking,” European Journal of Operational Research, 117, 199-210(1999).

2.Chaudhry, S. S., L. Salchenberger, and M. Beheshtian, “A small business inventory DSS: design, development, and implementation issues,” Computer & Operational Research, 23, 63-72 (1996).

3.Fausett, Laurene, Fundamentals of Neural Network, Prentice Hall International, 1994.

4.Hill, R. M., and M. J. Dominey, “Inventory policies for all-or-nothing demand process,” International Journal of Production Economics, 71, 365-371(2001).

5.Hsu, P. H., C. H. Wang, Joseph Z. Shyu and H.C. Yu, “A Litterman BVAR approach for production forecasting of technology industries,” Technological Forecasting & Social Change, Vol.70, pp.67-82, 2002.