應用倒傳遞類神經網路與EWMA統計製程管制於冷凍鏈溫度監控之研究

The Study of Applying BPN and EWMA SPC to Cold Chain Temperature Monitoring

邵奕誠(Yi-Cheng Shaw)

指導教授 ： 陳凱瀛

國立臺北科技大學/管理學院/工業工程與管理研究所/碩士(2009年)

https://doi.org/10.6841/NTUT.2009.00267

摘要

近年來由於都市化的發展，使得供應鏈伴演起更重要得角色，其中關於增進資料收集的效率、即時性與快速資料傳輸的特性更是廣為研究於供應鏈內。冷凍鏈為供應鏈管理中的一環，其中溫度監控在冷凍鏈內更是扮演著很重要的角色。本研究旨在統計製程管制圖與類神經網路之技術來記錄溫度的資料，應用EWMA統計製程品管來製作管製圖並監控溫度的變化以及資料的前處理，並且利用倒傳遞類神經網路於溫度趨勢之預測，當有失溫的情況就可即時通知控制中心並採取相對應的動作。最後製作一個結合倒傳遞類神經網路、統計製程品管圖於溫度監控之系統，並利用樂高 (LEGO) 積木來模擬實際冷凍鏈與展示系統之應用。

關鍵字

冷凍鏈 ； 製程監控 ； 指數加權移動平均 ； 倒傳遞類神經網路 ； 樂高

並列摘要

Recently, with the development of urbanization, the enhancement efficiency of contactless, real-time features and high data transmission rate in supply chain management are widely discussed. The cold chain is one part of the supply chain, and especially the temperature monitoring plays a vital role in cold chain system. In this paper, we are aimed to apply EWMA control chart and artificial neural network technologies to collect temperature data. The back-propagation neural network is used to predict temperature shifts and trend. EWMA control chart is adopted to monitor temperature variation. As there're something wrong appened, the control center of an enterprise can do some actions immediately to prevent further disaster. Finally, construct a system with back-propagation neural network, statistical process control chart. Simulations and demonstrations of real environment would be implemented by LEGO bricks at last.

並列關鍵字

Cold chain ； Process monitoring ； EWMA ； BPN ； LEGO

參考文獻

[80] Yu-Feng Huang, “Research of Applying RFID in Picking Process of Warehouse System.”, National Taipei University of Technology, 2007.

[66] Stephen B. Miles, Sanjay E. Sarma, and John R. Williams, “RFID Technology and Applications.”, Cambridge University Press, 2008.

[11] Chen-Hong Yang, “Process Monitor for Autocorrelated Data by Intrgrating ICA and Neural Network.”, Insitute of Commerce Automation and Management, National Taipei University of Technology, June, 2007.

[39] J. A. Anderson and E. Rosenfeld, “Neurocomputing: Foundations of Research.”, The MIT Press, December 18, 1989, ISBN 978-0262510486.

[1] A. A. E. Desouky and M. M. Eikateb, “Hybird adaptive techniques for electricload forecast using ANN and ARIMA.”, IEEE Proceeding Part C: Generation, Transmission and Distribution, Vol.147, 2000, pp.213-217.