半導體傳輸整合步進機在新產品/新製程導入情境下之排程

Abstract

中文摘要 本論文主要研究在半導體新產品/新製程的情境下，傳輸整合步進機的排程問題。傳輸整合步進機通常是半導體工廠中的瓶頸機台，它的內部結構是由一連串的反應室所組成，而其對外連接了一個有限埠(port)存放區。在傳輸整合步進機內部反應室的加工與傳輸單位為晶片，而其外部埠的傳輸單位為一生產批量(最大可包含 25 片晶片)。因為傳輸單位的不一致，導致傳輸整合步進機的產能損失，特別是會發生在新產品/新製程導入情境下。在此情境下機台需要生產許多小批量(低於每批 25 片晶片)，而這種產能損失可以透過有效的工件排程降低其損失。 本研究主要分成三個部分，第一部分在單部步進機的情境下，發展一個基因演算法求解其工件排程問題。第二部分為了得到更好的解題品質，發展新的演算法：GA-Tabu。第三部分探討在多機情境下，求解工件排程問題，在此情境下產生兩個決策變數：工件指派給各機台，以及工件在機台內的排序問題；為了解決此問題，本研究透過新的染色體設計，能同時代表此兩個決策變數，使得解題空間變小，求得更好的解題品質。 經過實例驗證顯示本研究新創的 GA-Tabu 演算法均超越過去文獻發展的演算法。

關鍵詞：排程、半導體、流程式生產、埠區限制、基因演算法、禁忌搜尋法

Abstract This dissertation examines job scheduling problems for in-line steppers operating in a new process/production introduction (NPI) scenario. An in-line stepper is a bottleneck machine in a semiconductor fab. Its interior is comprised of a series of chambers, while its exterior is a dock equipped with a limited number of ports. The transportation unit for each chamber is a piece of wafer while that for each port is a job that could contain up to 25 wafers. This transportation incompatibility may lead to an unexpected capacity loss for an in-line stepper—in particular in an NPI scenario that, by nature, includes a substantial number of small-sized jobs. Such a capacity loss can be alleviated by effective scheduling. This dissertation is composed of three parts. Firstly, we develop a genetic algorithm (GA), a kind of meta-heuristic algorithm, to solve the job scheduling problem for a single in-line stepper. Secondly, to enhance the solution quality, we develop some other meta-heuristic algorithms. Thirdly, we examine a job scheduling problem, in which jobs could be processed by multiple in-line steppers. This problem includes two decisions: job assignment and job sequencing. Through developing a new and concise representation scheme for modeling the two decisions, we solve the multiple-machine scheduling problem by various meta-heuristic algorithms. The preceding single-machine scheduling problem in fact is a special case of the multiple-machine scheduling problem. Extensive numerical experiments indicate that a GA-Tabu algorithm developed by us essentially outperforms the other meta-heuristic algorithms adopted from literature.

Keywords: Semiconductor, In-line stepper, Scheduling, GA-Tabu, Flow shop, Port constraint, GA