線材批量生產交期限制下之利潤最大餘料最少排程規劃

Abstract

冷陰極螢光燈管(CCFL, Cold Cathode Fluorescent Lamps)為薄膜電晶體液晶顯示器(TFT-LCD, Thin Film Transistor-Liquid Crystal Display)的上游零件，其扮演面板背光模組中提供光源的角色，國內冷陰極燈管所使用之電極大部分皆是由國外生產，如何提高國內電極製造廠商的競爭力，成了重要的課題。冷陰極燈管電極製造流程中使用材料大部分單位都使用Pcs，但除了 2 Parts Welds 此製程中的 DUX 線是採用公斤再轉換為Pcs，生產製造單位的不同造成生管於排程管理以及材料控管上遭遇很大的難題，又因為客戶規格之不同在轉換生產時造成過多的餘料浪費(Loss Ratio)以及設備設置時間浪費，本研究結合理論概念與實務特性，利用基因演算法進行最佳化生產作業規劃；就實際個案數據驗證分析結果來說，電極2 Parts Welds 製程可收斂到最佳生產批量，根據此一最佳生產批量進行生產排程規劃，將有效地增進CCFL生產獲利空間與加速CCFL製程效率，連帶提昇台灣的CCFL製造廠商的國際競爭力。

With the rapid development of the CCFL (Cold Cathode Fluorescent Lamps) technology and its related applications in Taiwan, the wide use of the CCFL is the industrial use in backlight of LCD (Liquid Crystal Display).The key component of CCFL is the cathode.In 2 Parts Welds process of the cathode, there are different measurements or production units between importable materials and exportable products , so the 2 Part Welds process of Cathode is a specific production scheduling issue . In terms of 2 Part Welds process of the CCFL, this paper proposes the special scheduling problem that aims to maximize the profit and simultaneously conform to the minimum of utilization rate in Dux wires. In nature, this paper proposed a Genetic Algorithm (GA) to solve this problem. Finally, the experimental results have shown that the GA can solve the proposed scheduling problem and find the efficiently and effectively bach size.