供應鏈管理下需求預測控制之研究

A Demand Predictive Control Model for Supply Chain Management

胡大瀛(Ta-Yin Hu) ； 蘇怡如(Yi-Ru Su)

《運輸學刊》 18卷4期 (2006/12) Pp. 417-446

https://doi.org/10.6383/JCIT.200612.0417

摘要

在全球化競爭的情況下，供應鏈管理(Supply Chain Management, SCM) 成為各企業在經營管理上的一個重要課題。本研究主要構建動態供應鏈管理模式，並將模式預測控制(Model Predictive Control, MPC)應用於供應鏈管理。首先定義供應鏈組成架構並分析架構階層間之關係及限制，依此構建供應鏈模式。接著構建模式預測控制，以滾動平面法(rolling horizon approach)結合過去及現在的控制行為預測未來之控制行為，在系統因素或環境有所變動時進行決策更新；此外，構建一自我迴歸移動平均整合(Auto Regressive Integrated Moving Average, ARIMA)時間序列模型進行需求預測。實驗案例設計上分為兩部份，首先建立需求預測模型，再進行最佳化求解；在需求預測模型之建立上，其實驗數據來源以實際訂單數列資料校估獲得預測模型參數。供應鏈所需資料則由不同型態之虛擬資料進行最佳化以獲得決策，以驗證模型。

關鍵字

供應鏈管理 ； 需求預測 ； 模式預測控制

並列摘要

In the global competition, Supply Chain Management (SCM) plays an important role in enterprise operations management. This research constructs a dynamic SCM model, and applies MPC to supply chain management. The SCM model is constructed based on the defined supply chain structure, and then elements in the SCM are analyzed to explore the relationship and restriction of variables among the echelon structure. In demand forecasting, an Auto Regressive Integrated Moving Average (ARIMA) time series model is implemented to forecast future demands based on historical sale records. The MPC includes decision variable and demand forecasting models. Based on the past and present control actions, a rolling horizon approach is proposed to predict future control strategies. In addition, the case study is divided into two parts, demand forecasting model and optimization model. The demand forecasting model is calibrated based on the actual data. The proposed supply chain system is verified by numerical analysis through a set of simulated data.

並列關鍵字

Supply chain management ； Demand forecasting ； Model predictive control

參考文獻

臺灣區自行車輸出業同業公會

蔡柏灃(2003)。類神經網路模式預測控制。清華大學化學工程研究所。

Akaike, H.(1974).Information Theory and an Extension of the Maximum Likelihood Principle.Proceedings of the 2nd International Symposium on Information Theory.(Proceedings of the 2nd International Symposium on Information Theory).

Braun, M. W.,Rivera, D. E.,Flores, M. E.,Carlyle, W. M.,Kempf, K. G.(2003).A Model Predictive Control Framework for Robust Management of Multi-product, Multi-echelon Demand Networks.Annual Reviews in Control.27(2),229-245.

Bose, S.,Pekny, J. F.(2000).A Model Predictive Framework for Planning and Scheduling Problems: A Case Study of Consumer Goods Supply Chain.Computers and Chemical Engineering.24(2),329-335.