| 摘要: | 近十年來半導體產業已是全世界非常重要的產業之一，在未來，物聯網、人工智慧與電動汽車等科技逐漸崛起，而半導體是這些未來科技的主要元件材料，所以不管是現在還是未來，世界半導體協會、台灣半導體產業協會或工研院等機構皆認為半導體產業都將蓬勃發展並帶動半導體的供應鏈，每年的市場成長將不斷的上升，其中台灣擁有全球最完整的IC產業鏈及上下游的專業分工，使台灣在2018年拿下IC製造與封測全球第一，而IC設計位居全球第二的好成績。
陶瓷基版為台灣IC產業下游中的其中一環，比一般基板更適用於較惡劣的環境，且性能穩定，其最大優點為耐熱，使它廣泛應用於功率電子、電子封裝、混合微電子與多晶片模塊等領域，國內外競爭廠商有Kyocera、Murata、同欣電子、禾伸堂、九豪和璦司柏等，本研究以電鍍銅製程為插單的主要探討，電鍍銅製程為陶瓷基板中的主要製程之一，製造時間長且產量需求龐大，此機台是屬於平行機台，在排程問題中是常見的問題類型之一，本研究使用混合整數規劃方法求解此站的緊急訂單排程問題
，以找出最小延遲成本的最佳解。
大多製造業都是顧客選擇公司，而不是公司考慮是否接單，尤其是重要客戶的緊急訂單，為了符合實際狀況，本文所考慮的是接受所有急單，為了讓決策者跳脫大客戶優先生產的舊有觀念，本研究依照決策者們的主觀認知給予所有訂單不同的權重，當發生訂單延遲時，普通單與急單會根據不同權重有不同的延遲成本，求出延遲成本後進而比較插單前後所帶來的影響，以便給決策者做適當的決策，經過小筆數據模型測試後，本研究以週排程為基準並且模擬實際例子，最後依照求解中所面臨的問題與求解的結果進行結論與建議。
;In the past ten years, the semiconductor industry has become one of the most important industries in the world. Such as the Internet of Things, artificial intelligence and electric vehicles will gradually emerge in the future. Semiconductors are the main component materials of these future technologies, so whether it is now or in the future, SEMI、TSIA or ITRI, and so on believed that the semiconductor industry will flourish and drive the supply chain of semiconductors. The demand of the market will continue to rise every years.
Among them, Taiwan has the most complete IC industry chain in the world and the professional division of work between upstream and downstream, which makes Taiwan become the number one in the world in IC manufacturing and testing in 2018. And IC design ranks second in the world.
Ceramic substrate is one of the downstream of IC industry in Taiwan. It is more suitable for harsh environment than general substrate, and its performance is stable. Its biggest advantage is heat tolerance, which makes it widely used in power electronics、electronic package、hybrid microelectronics and MCM. Domestic and foreign competitors include Kyocera、Murata、Tong Hsing、HolyStone、LEATEC Fine Ceramics and ICP. This study focuses on the rush order in the Plating Copper process. The Plating Copper process is one of the main processes in the Ceramic substrate. The manufacturing time is long and the production demand is huge. This machine belongs to the parallel machine and is one of the common types of problems in the scheduling problem. This study uses a mixed integer programming method to solve the rush order in scheduling problem at this station to find the best solution for the minimum delay cost.
In most manufacturing situations, the company is selected by the customer, not the company considering whether to take orders, especially the rush orders of important
customers. In order to meet the actual situation, our study considers accepting all the rush orders. In order to let decision maker think outside the box that important client not priority production. In this study, all the orders are given different weights according to the subjective of the decision makers. When an order delay occurs, the ordinary order and the rush order will have different delay costs according to different weights. After calculating the delay cost, we compare the effect of whether we want to have rush order or not, in order to make appropriate decisions for the decision makers. After the small data model test, this study takes the weekly schedule as the benchmark and simulates the actual situation. Finally, conclusions and suggestions are made according to the problems faced after the solution and the results of the solution. |
