臺灣博碩士論文加值系統

在半導體的製程中，指數加權移動平均 (Exponentially Weighted Moving Average, EWMA) 控制器是半導體製程中最常被使用於批次控制 (Run-to-Run control, R2R) 的手法之一。所謂EWMA回饋控制器是根據產出資料的先後順序給予不同的權重進行疊代，估計模型中之未知參數進而調整製程的投入變數。
在現今半導體產業中，存在成本的問題，大多屬於單一機台生產多樣產品或是多機台生產多產品，考慮到各產品之截距及斜率可能有所差異，無法直接採用single EWMA控制器來監控此生產製程。因此，針對此生產製程，文獻上已有學者提出Product-based控制器、Tool-based控制器、改良式的 tool-based 控制器、MTB 控制器(Modified tool-based, MTB 控制器)、轉換產品控制器及DREWMA控制器 (Distrubance retuned EWMA, DREWMA控制器)。這些控制方法雖可以逐步調整多產品到目標值的生產模式，但在產品交替生產的過程中，轉換產品後之製程產出值與目標值的偏差量卻無法有效地被降低，導致製程產出值的總均方差 (Total mean square error, TMSE) 較大，且針對不同的製程干擾，不一定適用於每個控制器，因此，本文採用模擬的方法，在最佳的折扣因子下，以改良式轉換產品控制器為基準，利用產品轉換後的前五期產出值的總均方差準則，來探討不同組合的製程干擾下，各種控制器的的相對效率，做優劣比較與分析。

Exponentially weighted moving average (EWMA) is commonly used in Run-to-run (R2R) control of the semiconductor manufacturing process. The EWMA feedback controller is a popular model-based controller which primarily uses data from past process run to adjust the input recipe for the next run. In the literatures, several papers addressed the efforts of long-term stability and short-term performance of EWMA control scheme. However, in practical applications, multiple-tool and multiple- product production style is common for the practical implementation of R2R control scheme. A single EWMAcontroller can not be directly used to monitor the production process. Thus, for a such mixing production style, product-based controller, tool-based controller, MTB controller, modified sing EWMA controller and DREWMA controller have been proposed. Therefore, in this paper, we adopt the Distrubance retuned EWMA (DREWMA) controller to discuss the improvement of total means square error (TMSE), to explore different combinations of process disturbances under various controllers and its relative efficiency and to make comparison and analysis of the merits and drawbacks so as to efficiently use the existing controller to significantly reduce TMSE.

第一章 簡介 1
1.1 前言 1
1.2 批次控制簡介 2
1.3 研究動機 4
1.4 研究架構 4
第二章 文獻回顧及問題描述 6
2.1 single EWMA控制器簡介 6
2.2 多產品製程控制器簡介 7
2.2.1 Product-based 控制器簡介 7
2.2.2 Tool-based 控制器簡介 11
2.2.3 轉換產品控制器簡介 13
2.2.4 改良式的 tool-based 控制器(Modified tool-based, MTB 控制器)簡介 16
2.2.5導入誤差項控制器(Distrubance retuned EWMA,DREWMA控制器) 17
2.3 問題描述 19
第三章 各控制器之績效表現 21
3.1 固定斜率初始估計量，改變截距初始估計量下，各控制器之績效表現 21
3.2 固定截距初始估計量，改變斜率初始估計量下，各控制器之績效表現 25
第四章 結論與後續研究 30
參考文獻 31

Ai, B., Zheng, Y., Wang, Y., Jang, S. S., Song, T. (2010), “Cycle forecasting EWMA (CF-EWMA) approach for drift and fault in mixed-product run-to-run process,” Journal of Process Control, 20, pp. 689-708.
Ingolfsson, A. and Sachs, E. (1993), “Stability and sensitivity of an EWMA controller,” Journal of Quality Technology, 25, pp. 271-287.
Tseng, S. T., Chou, R. J. and Lee, S. P. (2002), “Statistical design of double EWMA controller,” Applied Stochastic Models in Business and Industry, 18, pp. 313-322.
Tseng, S. T., Yeh, A. B., Tsung, F., Chan, Y. Y. (2003), “A study of variable EWMA controller,” IEEE Transactions on Semiconductor Manufacturing, 16, pp. 633-643.
Zheng, Y., Lin, Q. H., Wong, D. S. H., Jang, S. S., Hui, K. (2006), “Stability and performance analysis of mixed product run-to-run control,” Journal of Process Control, 16, pp. 431-443.
Zheng, Y., Ai, B., Wong, D. S. H., Jang, S. S., Wang, Y., Zhang, J. (2010), “An EWMA algorithm with a cycled resetting (CR) discount factor for drift and fault of high-mix run-to-run control,” IEEE Transactions on Industrial Informatics, 6, pp. 229-242.
簡勁舟 (2011),“單機台多產品批次回饋系統之績效分析,”國立清華大學統計研究所碩士論文.
李家豪 (2012),“多產品混合生產製程之批次回饋控制分析,”國立清華大學統計研究所碩士論文.
王昱凱 (2013),“導入誤差項多產品製程之批次控制績效分析,”靜宜大學財務與計算數學所碩士論文.