指數平滑與移動平均法在回收率不確定下再製生產長鞭效應之探討

Investigating the Bullwhip Effects of Remanufacturing with the Uncertainties of Return Rate by Using the Forecasting Methods of Exponential Smoothing and Moving Average

張家翊(Chia-I Chang)

指導教授 ： 邱裕方

中原大學/電機資訊學院/工業與系統工程研究所/碩士(2009年)

https://doi.org/10.6840/cycu200901118

摘要

在過去數十年中，電子與資訊產品帶領人類邁入便利與快捷的新時代，但同時也產生了巨量的資訊產品廢棄物，使環境汙染的問題日益嚴重。為避免環境繼續遭受破壞，世界各國的環保意識逐漸升高，紛紛投入資源於再生利用產品的開發，使企業能兼顧環境保護與節省自然資源的義務。逆向物流是將舊產品從消費者處回收、再處理、再製造、再重新銷售到市場上的過程，多年來學者們已相繼發展有效的逆向物流成本模式。由於逆物流與再製造是近幾年發展的研究，早期的逆向物流多數研究需求固定下的成本問題，卻鮮少研究因為「回收率變動」所導致存貨與長鞭效應變異增大的問題。本研究延伸正向物流的觀念，導入回收產品與運輸成本的因素，計算零售商最適存貨，並以最佳存貨估計零售商每期最高訂購點，再以移動平均法與指數平滑法預測顧客需求，探討不同參數下，使用移動平均法與指數平滑法預測顧客需求值。最後透過統計假設檢定驗證在逆向物流裡這兩種預測方法有顯著的差異，其中移動平均法能使兩階段供應鏈裡的長鞭效應降低。

關鍵字

逆向物流 ； 長鞭效應 ； 指數平滑法 ； 移動平均法 ； 存貨

並列摘要

In last few decades, electronic and hi-tech products lead people to a new era, but it also produced a huge amount of waste of technological products, so the problems of environmental pollution are getting worse. In order to avoid the continuous damage on environment, countries around the world began to invest in resources for the development of reused products due to the sense of environmental protection progressively, and enterprises can take into account environmental protection and saving natural resources obligations. Reverse logistics is a process of return, re-processing, re-manufacturing and re-sold that let old products from consumers to the markets. Over the years, scholars developed some effective cost models of reverse logistics, so enterprises can increase operating income by saving costs. Due to the research of reverse logistics and remanufacturing developed in these years, most researches of reverse logistics investigated the cost problem of fixed demand, but few of them considered the problem of increasing variation of inventories and bullwhip effect caused by the changes of return rate. This research considers the recovery products and transportation cost factors to calculate the optimal inventory for retailers to extend the forward logistics models, and estimate the order-up-to point in each period of retailer, then forecasting the customer demand by the method of moving average and exponential smoothing to investigate the effects of bullwhip effect under different parameters. This research found that there are significant difference between the two forecasting methods, and the moving average forecasting method can decrease the bullwhip effect in two-echelon supply chain by statistical testing hypothesis.

並列關鍵字

Moving average method ； Exponential smoothing method. ； Bullwhip Effect ； Reverse Logistics ； Inventory

參考文獻

3. Agrawal, S., R. N. Sengupta and K. Shanker (2007), “Impact of information sharing and lead time on bullwhip effect and on-hand inventory,” European Journal of Operational Research, Volume 192, Issue 2, Pages 576-593.

4. Alshamrani, A., K.Mathur and R. H. Ballou (2007), “Reverse logistics: simultaneous design of delivery routes and returns strategies,” Computers & Operations Research, Volume 34, Issue 2, Pages 595-619.

6. Biehl, M., E. Prater and M. J. Realff (2007), “Assessing performance and uncertainty in developing carpet reverse logistics systems,” Computers & Operations Research, Volume 34, Issue 2, Pages 443-463.

10. Brito, M. P. D. and E. A. van der Laan (2007), “Inventory control with product returns: The impact of imperfect information,” European Journal of Operational Research, Volume 194, Pages 85-101.

11. Caloiero, G.., F. Strozzi and J.Zaldivar Comenges (2008), “A supply chain as a series of filters or amplifiers of the bullwhip effect,” Int. J. Production Economics, Volume 114, Pages 631-645.