製造程序常會遭受到一些或大或小的特殊干擾，像是漂移或速移。在很多情況下，若我們可以找出製程受到干擾的原因，但此原因無法消除或消除所需的成本過高，在這樣的情況下，當製程產出的偏移可以藉由對機台設定值的調整而加以彌補時，我們即可以使用批次回饋控制技術。
在本論文中，我們提出一個能控制隨機性速移與漂移干擾之製程批次整合控制器。針對製程的漂移干擾，我們以雙加權指數移動平均(Double Exponentially Weighted Moving Average; D-EWMA)的原理設計了D-EWMA控制器。同時，在增加自我調整(Self-tuning; ST)的機制到D-EWMA控制器後，我們亦發展了能同時有效控制隨機性速移干擾的STD-EWMA批次整合控制器。
利用蒙地卡羅(Monte Carlo)的模擬結果，我們驗證了本論文所提出的STD-EWMA批次整合控制器的優越性，當製程面對隨機性的速移或漂移干擾時，STD-EWMA控制器均能有效控制製程干擾，並且其控制績效均優於現行的批次控制方法，諸如EWMA控制器或是PCC控制器。

Manufacturing processes are subject to small and large special disturbances such as process drifts or shifts. In many cases the causes of disturbances are known, but it is either impossible or too expensive to remove them. In such cases, when the resulting output deviations can be compensated by adjusting the processing recipe, process control techniques such as the run-by-run feedback control scheme will be useful.
In this thesis, we present an integrated run-by-run controller for processes subject to random shifts and drifts. For controlling process drifts, the controller has been designed based on the formulation of Double Exponentially Weighted Moving Average (D-EWMA). By adding the self-tuning (ST) feature into the D-EWMA controller, the resulting STD-EWMA controller is able to controlling random shifts more effectively.
Monte Carlo simulation results validate that our STD-EWMA controller is superior to the current run-by-run process controllers such as the EWMA controller or Predictor-Corrector controller.

論文摘要I
目錄III
表格目錄IV
圖表目錄V
第1章 緒論1
1.1 研究動機1
1.2 研究目的3
1.3 研究限制4
1.4 論文架構4
第2章 文獻探討6
2.1 文獻探討6
2.2 控制器簡介7
2.2.1 EWMA控制器7
2.2.2 PCC控制器11
第3章 D-EWMA控制器15
3.1 EWMA控制器的特性分析15
3.2 PCC控制器的特性分析18
3.3 D-EWMA控制器的特性分析23
3.4 模擬驗證30
第4章 STD-EWMA控制器42
4.1 ST-EWMA控制器42
4.2 STD-EWMA控制器46
4.3 模擬驗證51
第5章 結論與未來研究建議68
5.1 結論68
5.2 未來研究建議68
參考文獻70

[1] G. Box and M. Jenkins, Time Series Analysis - Forecasting
and Control. Oakland, CA:Holden-Day, 1974.
[2] E. Sachs, A. Hu, A. Ingolfsson, “Run by Run process
Control: Combining SPC and Feedback control,” IEEE
Transactions on Semiconductor Manufacturing, vol 8,
Feburuary. 1995.
[3] J. Stefani, S. Poarch, S. Saxena, and P. Mozumder,
“Advanced Process Control of a CVD Tungsten Reactor,”
IEEE Transactions on Semiconductor Manufacturing, vol. 9,
no. 3, August 1996.
[4] J. Moyne, N. Chaudhry, and R. Telfeyan, “Multi-Branch Run-
to-Run Controller for Plasma Etching,” Journal of Vacuum
Science and Technology, 1995.
[5] T. Smith, D. Boning, “A Self-Tuning EWMA Controller
Utilizing Artificial Neural Network Function Approximation
Techniques,” IEEE Transactions on Components, Packaging,
and Manufacturing Technology, March 1997.
[6] R. Guo, L. Huang, A. Chen, and J. Chen, “A Cost-effective
Methodology for a Run-by-Run EWMA Controller,”
International Symposium on Semiconductor Manufacturing,
October 1997.
[7] 黃麗霞,”結合統計製程管制與回饋控制之非連續批次控制器－以半
導體製程為例,”台大商研所碩士論文,1996.
[8] 郭瑞祥, 陳進忠, 黃麗霞, 王聖雅 “半導體製造批次製程控制”,
1997自動控制研討會
[9] B. J. Mandel, “The Regression Control Chart,” Journal of
Quality Technology, Vol.1, No.1, January. 1969.
[10] J. Hunter, “ The Exponential Weighted Moving Average,”
Journal of Quality Technology, vol.18, no. 4, October. 1986
[11] S. W. Roberts, “Comtrol Chart Tests Based on Geometric
Moving Average,” Technometrics, vol.1, no.3, August 1959.
[12] M. Klein, “Composite Shewhart-EWMA Statistical Control
Schemes” IIE Transactions, vol.28, June 1996.
[13] Douglas C. Montgomery, Introduction to Statistical Quality
Control, John Wiley & Sons, Inc., New York, NY. 1996.
[14] A. Ingolfsson,. and E. Sachs, “Stability and Sensitivity
of an EWMA Controller,” Journal of Quality Technology,
vol. 25, no. 4, October 1993.
[15] J. Chen , R. Guo, A. Chen and S. Lu “A Self-Tuning EWMA
Controller for Processes subject to Small and Large
Disturbances,” International Symposium on Semiconductor
Manufacturing, Tokyo, Japan, 1998
[16] S. W. Butler and J. Stefani, “ Supervisory Run-to-Run
Control of Polysilicon Gate Etch Using In Situ
Ellipsometry,” IEEE Transactions on Semiconductor
Manufacturing, vol. 7, no. 2, May 1994
[17] T. Smith, D. Boning, J. Stefani, and S.Butler, “Run by
Run Advanced Process Control of Metal Sputter Deposition,”
IEEE Transactions on Semiconductor Manufacturing, vol.11,
no.2 May 1998
[18] E. D. Castillo, “A Multivariate Self-Tuning Controller
for Run-to-Run Process Control under Shift and Trend
Disturbances,” IIE Transactions, vol. 28, 1996.
[19] E.D. Castillo, and A. Hurwitz, “Run-to-Run Process
Control: Literature Review and Extensions,” Journal of
Quality Technology, vol. 29, no. 2, April 1997
[20] E. D. Castillo, “On Properties of EWMA Run-to-Run
Feedback Controllers,” Industrial Engineering Research
Conference, May 1999
[21] R. G. Brown, “Smoothing, Forecasting and Prediction of
Discrete Time Series,” Prentice-Hall, Englewood Cliffs,
N.J.,1963
[22] C. C. Holt, “Forecasting Seasonals and Treand by
Exponentially Weighted Moving Average,” ONR Research
Memorandum, no. 52, April, 1957