| Abstract: | 光電產業為我國目前兩兆雙星計畫中的重點產業，其重要性極為顯著，其中又以液晶薄膜顯示器(TFT-LCD)產業為大家注目的焦點。目前我國與韓國已有研究使用模擬技術與經驗法則解決TFT-LCD產業複雜的單廠排程問題，但是對於TFT-LCD產業中重要的多階多廠區(Multi-Level,Multi-Site;MLMS)生產規劃問題鮮少有完整的論述，其主要原因為TFT-LCD在生產階層(stages)間具有相當多不同的特性與要求，如等級互用、等級組合、廠間物料共用等特性，致使TFT-LCD產業之多階多廠區規劃不同於以往較常見的組裝業供應鏈規劃。有鑑於此，本研究將探討TFT-LCD產業多階多廠區生產規劃模式，以作為業界與學界參考之用途。本研究將分兩年進行，第一年探討TFT-LCD產業多階多廠區生產規劃的特性與規劃模式，並設計生產規劃中「產銷平衡」功能的解決方法；第二年設計生產規劃中ATP/CTP功能與訂單滿足功能的解決方法，並實作與驗證本研究提出之多階多廠區生產規劃模式。整體而言，本研究之目的可分為下列幾點:1.由供應鏈規劃文獻與TFT-LCD產業特性，定義出台灣TFT-LCD多階多廠區規劃的定位與數學模式。2.發展TFT-LCD產業多階多廠區生產規劃中『產銷平衡』、『ATP/CTP』與『顧客訂單滿足』等三個最重要的功能模式。3.提供台灣TFT-LCD業界在多階多廠區生產規劃上的參考規劃模式，並使學界熟悉TFT-LCD產業多階多廠區規劃模式。在計畫的第一年中,主要任務為提出TFT-LCD多階多廠區生產規劃問題的特性與定位，並且發展TFT-LCD產業多階多廠區生產規劃中『產銷平衡』的演算方式。
TFT-LCD (Thin Film Transistor, Liquid Crystal Display) industry is one of the two most important industry recognized by the government in Taiwan. Currently, some researches have studied the production scheduling problem of single-site using simulation and heuristic techniques, but multi-level, multi-site (MLMS) production planning problem of TFT-LCD industry has not been widely studied due to the distinct production characteristics among different stages (e.g., grade classification and material sharing characteristics in TFT, LCD, LCM factories). The objective of this research is therefore to develop a TFT-LCD multi-level, multi-site production planning model which may provide a reference for academy and industry. This research will last for two years, the first year will study the characteristics and model of TFT-LCD multi-level, multi-site production planning, and develop the supply-demand balancing function model; the second year will develop both the ATP/CTP and order fulfillment function models, and evaluate the performance of the constraint-based genetic algorithm multi-stage, multi-site production planning model developed in this research. 2 The expected research results may be summarized as follows: (1) Characteristics of a TFT-LCD industry's multi-level, multi-site production planning function (1st year). (2) A TFT-LCD industry's multi-level, multi-site production planning mathematical model (1st year). (3) A constraint-based genetic algorithm multi-level, multi-site production planning model which includes (a) supply-demand balancing function model, (b) ATP/CTP function models, and (c) order fulfillment function model. (4) Implementation and evaluation of the performance of the proposed TFT-LCD industry's multi-level, multi-site production planning system (2nd year). |
