- 學位論文
應用費波南西系數改善限制理論下之補貨管理
The application of Fibonacci Coefficient to improve the theory of Constrant in inventory management
摘要
若要品評ㄧ間企業或工廠績效的好與壞，存貨管理無非是重要指標其中之ㄧ。為了保證有足夠的量已供應下游的需求變動，存補貨管理在供應鏈中扮演著舉足輕重的角色。與傳統補貨策略相比較，限制理論(Theory of Constraints, TOC)所主張的需求拉動補貨(Demand-Pull)結合緩衝管理(Buffer Management)為ㄧ套有效的供應鏈存貨管理方法(有許多文獻證明) 。本研究則在其中加入了指數加權移動平均法(Exponentially Weighted Moving Average, EWMA)、安全庫存中的再訂購點(Reorder Point, ROP)、歷史波動率(History Volatility, HV)、費波南西數列(Fibonacci Number)和由費波南西所延伸出的黃金切割率(Golden Ratio)並加以改良成適合計算庫存相關的模式，藉此達到只需考量顧客即時需求數據和需求的趨勢便能做出最適當的補貨策略。 本研究所評估的商品為半導體封裝產業的產品，其商品特點主要有生命週期短和需求變動大，故使用指數加權移動平均法(Exponentially Weighted Moving Average, EWMA)以整合顧客之實際需求量(Demand)與實際需求預測(Forecast)並加入再訂購點(Reorder Point, ROP)的算法以輔助實際需求量與實際需求預測可能忽略的小細節。在緩衝管理中將其比例依黃金切割率(Golden Ratio)為基礎將其分乘六種不同的模式，利用歷史波動率(History Volatility, HV)來計算廠商所給之實際需求預測，以辨別該商品之波動性，明確的指出該商品所適用的補貨模式。
並列摘要
To assess the efficiency of an enterprise or a factory, without a doubt inventory management is one of the important indicators. To ensure a sufficient quantity to supply the downstream demand changes, the management of stockpiles is important in the supply chain. Compared to the traditional replenishment strategy, Demand-Pull and Buffer Management advocated by Theory of Constraints (TOC), is an effective set of supply chain inventory management method (proved by many references). This method includes an Exponentially Weighted Moving Average (EWMA), Reorder Point (ROP) in the safety stock, History Volatility (HV), Fibonacci series (Fibonacci number), and the extension of Fabian Nancy’s Golden Ratio (Golden Ratio) to gain a suitable inventory-related model. The objective is to make the most appropriate replenishment strategy by considering customer demand data and the demand trends. The products evaluated in this project are from a semiconductor packaging industry. Its products are characterized mainly by short life cycle and large changes in demand. Therefore, the EWMA is used to integrate the Customer’s actual demand and the actual demand forecasts and adds ROP algorithms to assist in small details that may be overlooked. For the buffer management, we divide the ratio into six different modes based on the Golden Ratio. The Historical Volatility is used to calculate the actual demand forecast given by the manufacturer to identify the volatility of the commodity and to clearly indicate the replenishment model to which the commodity is applicable.