應用貝氏管制圖與追蹤訊號於需求監控存貨管理系統之探討

A Study on the Application of Bayesian Control Chart and Tracking Signal for Demand Monitoring and Inventory Management System

吳致賢(Chih-Hsien Wu)

指導教授 ： 蔣明晃

國立臺灣大學/管理學院/商學研究所/碩士(2017年)

https://doi.org/10.6342/NTU201701235

摘要

在全球化與資訊科技進步之下，使得市場資訊變得更加透明，加快了產品或服務相關資訊的傳遞速度，市場競爭變得更加激烈，因此，企業也不斷的試圖提高本身供應鏈的效率來降低成本，而存貨管理一直是供應鏈管理當中極為重要的議題。Watt et al. (1994) 首先提出將統計製程管理 (statistical process control, SPC) 中的管制圖 (control chart) 應用於存貨管理的概念，而後續也有多位學者將各式各樣的管制圖應用在存貨管理，但其多屬於頻率學派的管制圖與存貨系統之結合。本研究首先引進貝氏管制圖與追蹤訊號的概念於需求監控上，當有大量且即時性的資料時，貝氏管制圖能迅速進行更新，協助企業做出更快速、準確的決策。
本研究透過模擬不同需求變化的方式，與傳統定期盤存制存貨管理系統進行比較。結果發現，本研究所提出的需求監控存貨管理系統在需求穩定的情況之下，與傳統定期盤存制存貨系統在總成本上的表現沒有顯著差異，且當需求的波動變大時，會導致總成本微微上升，但可以發現其庫存水準比傳統的定期盤存制存貨管理低許多，可知本研究所提之系統可以有效的監控需求變化，減少不必要的庫存。而當需求遞增時，本研究提出之系統比傳統的定期盤存制存貨管理系統不僅減少了約一半左右的成本，同時可以提供較高的服務水準。

關鍵字

存貨管理 ； 貝氏管制圖 ； 追蹤訊號

並列摘要

With the globalization and rapid development of information technologies, market information has become more transparent, accelerating the delivery of information about products or services and making competition more intense. As a result, companies are constantly trying to improve the efficiency of their supply chain to reduce costs. Inventory management has always been an extremely important issue in supply chain management. Watt et al. (1994) first proposed the concept of the application of control charts in statistical process control (SPC) for inventory management. After that, a number of scholars had proposed and combined a variety of control charts with inventory management, but most of them belong to the frequency school. This study first introduces the concept of Bayesian control chart and tracking signal on demand monitoring. When there is a lot of real-time information, the Bayesian control chart can be updated quickly to help companies make faster and more accurate decisions.
This study simulates various patterns of demand and compares the system proposed in this study with the traditional (R, S) inventory management system. This study indicates that when the demand is constant, there is no any significant difference in total costs between the system proposed and traditional (R, S) system. Although the total cost will increase slightly when the demand fluctuates more intense, the inventory level of the system proposed is much lower than that of traditional (R, S) system, which means the system proposed can monitor the demand effectively. When the demand increases, the total cost of the system proposed is half of that of the traditional (R, S) system, and it can remain a higher service level simultaneously.

並列關鍵字

Inventory Management ； Bayesian Control Chart ； Tracking Signal

參考文獻

Aburto, Luis, & Weber, Richard. (2007). Improved supply chain management based on hybrid demand forecasts. Applied Soft Computing, 7(1), 136-144.

Aggarwal, Sumer C. (1974). A review of current inventory theory and its applications. International Journal of Production Research, 12(4), 443-482.

Alon, Ilan, Qi, Min, & Sadowski, Robert J. (2001). Forecasting aggregate retail sales: a comparison of artificial neural networks and traditional methods. Journal of Retailing and Consumer Services, 8(3), 147-156.

Carbonneau, Real, Laframboise, Kevin, & Vahidov, Rustam. (2008). Application of machine learning techniques for supply chain demand forecasting. European Journal of Operational Research, 184(3), 1140-1154.

Cheng, Jui-Chang, & Chou, Chao-Yu. (2008). A real-time inventory decision system using Western Electric run rules and ARMA control chart. Expert Systems with Applications, 35(3), 755-761.