運用模糊認知圖進行以RFID為基之逆物流模式決策支援方法與應用

Using Fuzzy Cognitive Map for RFID-based Reverse Logistic Model Decision Support and Application

吳昌儒

指導教授 ： 張瑞芬

國立清華大學/工學院/工業工程與工程管理學系/碩士(2009年)

https://doi.org/10.6843/NTHU.2009.00101

摘要

由於環保意識抬頭，以及相關的法令的制定(如：IEEE、RoHS……等)，自由的退貨政策，使得針對產品或零組件如何有效的回收、再利用成為近年來重要的議題之ㄧ。一般供應鏈主要是從供應者的供貨到製造者的加工製造，再到物流中心，再到零售商，最後才到消費者手上。而當我們增加環保的考量(例如：ISO14000 的認證)，或是其他的因素例如：提昇企業的形象進而增加銷售量、有效的資源利用，發展新方法降低製造成本和時間，延長產品的生命週期，和降低部分設備的設置等因素。企業可能就須在物流體系中再扮演廢物清理、回收再利用、再處理、再製造、再修理、退貨回收、的角色。以達到前述的效益，而關於這樣的流程我們稱它為逆物流(Reverse Logistics)。本研究希望透過建立整體供應鏈的模糊認知圖，對於整體逆物流的系統發生結合正向物流之整體供應鏈中各角色內部作業因子的影響評估和找出適當的決策因子狀態進行系統的調整，模糊認知圖能夠透過圖像化的方式讓決策者輕易的了解供應鏈當中各因子間的因果關係和交互影響程度，將即時收集的數據運用基因演算法的方式做為認知圖中權重的訓練，提供一個更客觀的逆物流之模糊認知圖模型。由於產品跟部分零組件的回收，在數量、地點、時間上較分散且不確定，因此，利用RFID的技術的導入，應用EPCglobal Network的架構下，進行有效率的監控供應鏈上產品/零組件/包裝容器流通以及相關的作業資訊。透過RFID的技術可解決在產品流通資訊上無法及時取得的問題，運用即時監控的特性可進行供應鏈上相關數據的收集，藉由本研究的模糊認知圖和資訊系統，結合了EPCglobal Network 和RFID 技術，再利用案例中的物流箱回收管理進行本研究所提出的推論與決策分析方法。透過這項研究使得公司企業在進行決策時，能夠達到迅速且有效的決策支援，提升競爭力和跟供應鏈的效率。

關鍵字

逆物流 ； RFID ； EPCglobal Network ； 模糊認知圖 ； 基因演算法

並列摘要

With increasing environmental consciousness, green directives (e.g., WEEE) and liberal return policies, recycling, re-use, re-assembly of materials, components and products are attracting more attention from manufacturers and the public. Consequently, reverse logistics research is being used to analyze the processes associated with the flows of products, components and materials from end users to re-users (e.g., second markets or landfills). The components may be widely dispersed during reverse logistics, which makes it difficult to efficiently collect, re-use and re-assemble disposed components for reprocessing and remanufacturing. As a result, Radio Frequency Identification (RFID) technology combined with EPCglobal Network architecture is applied to enable efficient product and component data collection and data transmission. The proposed information system uses RFID for real time data tracking and the EPCglobal network architecture defines the layers (RFID tags, RFID readers, EPC middleware, EPC information service, Object naming service, and EPC discovery service) for data transmission. This research develops a decision support model which integrates fuzzy cognitive maps and genetic algorithm. The major advantage of using cognitive maps is that the model and the relationships among nodes (states) are linguistically expressed both quantitatively and qualitatively. Combining the model and the information system, and integrating the EPCglobal network architecture and RFID technology, the goals of the research are achieved. Inference analysis contributes to the system response forecasting and decision analysis supports the manager make a stable reverse logistics system by adjusting some internal operation factors. This research provides a more comprehensive view of the supply chain, and achieves rapid and effective decision-making support, to enhance competitiveness and efficiency of the supply chain.

並列關鍵字

Reverse Logistics ； RFID ； EPCglobal Network ； Fuzzy Cognitive Map(FCM) ； Genetic Algorithm

參考文獻

[1] Athanasios, K.T., Ilias, K., Konstantinos, G.M., 2003, “Using fuzzy cognitive maps as a decision support system for political decisions,” Lecture Notes in Computer Science, Vol. 2563, pp. 172-182.

[3] Dickerson J.A., Kosko B., 1993, “Virtual worlds as fuzzy cognitive maps,” in Proc. IEEE Virtual Reality Annu. Int. Symp., pp. 417-477.

[4] Domdouzis, K., Kumar, B., Anumba, C., 2007, “Radio-Frequency Identification (RFID) applications: A brief introduction,” Advanced Engineering Informatics, Vol. 21, No. 4, pp. 350-355.

[5] Eden, C., Ackermann, F. 2004, “Analyzing cognitive maps to help structure issues or problems,” European Journal of Operational Research, Vol.159, pp. 673-686.

[7] Hagiwara M., 1992, “Extended fuzzy cognitive maps,” in Proc. 1st IEEE Int. Conf. Fuzzy Syst., pp. 795–801.