預測補貨策略對供應商成本之影響評估

Evaluating the impact of the forecasting and replenishing strategies on the cost of suppliers

蔡志宏(Chih-Hung Tsai)

指導教授 ： 林則孟

國立清華大學/工學院/工業工程與工程管理學系/碩士(2009年)

摘要

本研究探討在供應商管理存貨的二階存貨系統中，評估在工廠供給限制下的混合預測補貨策略對供應商成本之影響，以期望在快速供貨環境下，能持續降低供應商於HUB倉發生的存缺貨成本。所考量的範圍介於供應商工廠、中間HUB倉庫與客戶產生需求預測之間，涉及需求型態與變異、預測方法、補貨策略、工廠存貨配置決策與工廠供給限制。利用模擬方法來建構模式，並擷取各決策組合下對供應商HUB倉成本的量測績效，再使用多因子實驗設計手法，企圖尋找低成本效益的一般性建議結論。由實驗結果得知，預測方法與補貨策略確實存在顯著性交互作用關係，建議供應商先評估客戶提供的預測與歷史需求間的預測偏差，再依據屬於不同預測偏差特性參考量表配合對應的補貨策略，將能達到較低的成本效益。相對的，在工廠供給端，當工廠成品存貨有限而無法滿足多方需求時，對於有限存貨配置不可偏廢一方，應設法使得所有下游HUB倉庫皆能被部份滿足，例如採用平均分配存貨的配置法則，來克服有限存貨分配問題對成本的衝擊。另一方面，在工廠端的供給設置量應盡可能相依實際客戶需求水準作配置，同時可降低補貨策略造成長鞭效應需求放大的現象，過多或過少的工廠供給配置量皆會造成供應商成本的增加。因此，本研究分別對供應商管理存貨的需求端與供給端，提供供應商在預測補貨、工廠供給限制、存貨配置決策上的參數與決策組合建議，以協助供應商持續降低成本。

關鍵字

供應商管理存貨 ； 預測方法 ； 補貨策略 ； 目標庫存

並列摘要

The present study discusses the two-stage inventory system within Vendor Managed Inventory and evaluates the effect of hybrid forecasting and replenising strategies towards the supplier’s cost under the plant supply constraint. It is anticipated that the supplier’s cost of stock and outstock at HUB warehouse can be gradually decreased within the fast supply environment. Our concerns fall amongst the supplier’s plant, middle HUB warehouse, and the forecast of the customer’s demand, involving the areas of demand pattern and variation, forecasting method, replenishing strategy, inventory allocation decision, and plant supply constraint. By adopting simulation methodology, we establish a model. Moreover, we attempt to seek a general suggestive conclusion by selecting the best performance for the supplier’s HUB warehouse cost from various strategy combinations along with the multi-factor experimental design. Based on the results, there is indeed a significant difference between the forecasting method and the replenishing strategies. It is advised that the supplier firstly evaluate the forecasting bias between the forecast and demand-supply history offered by customers before adopting the corresponding replenishing strategy. By doing so, it is hoped to benefit the lower- cost circumstance. On one hand, when a plant has limited goods in stock that cannot meet multiple demands, it should strike a balance regarding the constrained stock allocation. The plant is supposed to meet the partial demand of all the downstream HUB warehouses. For instance, by applying the rule of equal stock allocation can overcome the negative impact resulted from the finite stock allocation to the cost. On the other end, the amount of stock allocated in plant should be in harmony with the customer’s demand, which can diminish the Bullwhip Effect caused by the replenishing strategy. Moreover, the allocated amount in plant, whether over or below the standard, would affect the increase of the supplier’s cost. Therefore, the present study offers suggestions about parameters and decisions combination in terms of replenishing strategy, stock from plant supply constraint, and inventory allocation decision for the supplier’s supply and forecasting for customer’s demand end, with a purpose to lower
the cost gradually.

並列關鍵字

Vendor Managed Inventory ； Forecasting methods ； Replenising strategies ； Target stocks

參考文獻

1. 張怡筠，「CPFR中滾動式預測的關鍵因子探討」，國立清華大學資訊系統與應用所碩士論文，2006。

2. 陳昭廷，「CPFR協同預測中之承諾機制的影響因子分析」，國立清華大學工業工程管理學所碩士論文，2007。

3. 謝依潔，「供應鏈滾動預測與訂單策略之模擬研究」，國立中央大學資訊管理學系碩士論文，2007。

4. 李建勳，「供應商管理庫存下之預測補貨策略」，國立清華大學工業工程管理學所碩士論文，2008。

5. Alexander, E. E., J. C. Taylor and P. J. Daugherty. “Automatic replenishment programs and level of involvement: performance implications”, International Journal of Logistics Management, 10(1), pp.25-36, 1999