本研究著重於發展統計程序控制(SPC)與自動程序控制(APC)於連續製程之整合控制策略。統計程序控制利用控制圖技術監視程序之異常操作(例如均值偏移)，進而發出警報；自動程序控制對程序輸入作持續調整去維持程序輸出在控制目標值。為了有效增強自動程序控制的效果，連續製程的輸出取樣間隔必須夠小，這使得輸出量測數據的相關性受程序及雜訊動態的影響而變大，增加了統計程序控制應用的困難。吾人綜合兩者之特點去設計有效之整合控制策略。
傳統的SPC控制圖技術假設輸出量測數據的不相關性，當其用於相關程序數據時就容易產生誤判警報。吾人發現一些改良式控制圖可有效應用於相關程序數據去偵測程序均值偏移之發生，另外適當放大取樣週期可降低程序輸出量測之相關性，使得SPC技術的應用更加有效。在APC方面，吾人利用最小方差控制將程序輸出變異度最小化，同時降低輸出數據的相關性，但其缺點為無法完全消除擾動的均值偏移。內部模式控制可完全消除該均值偏移，但其改善輸出變異度和降低輸出數據相關性的能力不如最小方差控制。整合控制策略即利用改良式SPC控制圖技術來監視程序操作，並即時調整自動控制模式。當程序處於正常操作時，使用最小方差控制以降低程序輸出的變異度，並提高SPC技術之準確性。當SPC監視系統發出程序異常警報後，自動控制模式改為內部模式控制以排除均值偏移。當均值偏移排除後，自動控制模式再改回最小方差控制。

This research emphasizes on the development of the integrated control strategy of statistical process control (SPC) and automatic process control (APC) for continuous processes. Statistical process control utilizes control chart techniques to monitor the abnormal operation of the process (e.g., a shift in its mean), and gives the alarm; automatic process control maintains the process output at the control target by adjusting the process input. To enhance the effectiveness of automatic process control, the output sampling period of a continuous process must be sufficiently large. However, this increases the correlation of the output measurement data because of the influence of the process and noise dynamics. As a result, the application of SPC techniques becomes more difficult. We combine the two control modes to design the integrated control strategy.
Traditional SPC control chart techniques assume that the output measurement data is uncorrelated. Otherwise, these techniques are likely to cause false alarms. We find a few improved SPC techniques that can effectively detect the occurrence of abnormal operation from correlated output data. In addition, a larger sampling period can reduce the correlation of the data, thus rendering the application of SPC techniques more effective. In the respect of APC, we utilize minimum variance control to minimize the process output variation and to reduce the correlation of process data. However, minimum variance control fails to eliminate the shift in the process mean. Internal model control is able to eliminate the shift completely, but it is inferior to minimum variance control in reducing the output variation and the data correlation. The integrated control strategy uses improved SPC control chart techniques to monitor process operation, and select the best automatic control mode in real time. As the process is in normal operation, apply minimum variance control to reduce the output variance and to enhance the accuracy of SPC techniques. As the SPC system gives the alarm about abnormal operation, the APC mode should be changed to internal model control to eliminate the shift. After the shift is eliminated, the APC mode should be changed back to minimum variance control.

[1] Bergh, L. G. and J. F. MacGregor, “Constrained Minimum Variance
Controllers: Internal Model Structure and Robustness Properties”,
Industrial and Engineering Chemistry Research, 26, 1558-1564, 1987.

[2] Box, G. E. P., G. M. Jenkins, and G. C. Reinsel, Time Series Analysis,
Forecasting and Control, 3rd edition, Prentice-Hall, New York, 1994.

[3] Box, G. E. P. and T. Kramer, “Statistical Process Monitoring and Feedback
Adjustment – A Discussion”, Technometrics, 34, 251-257, 1992.

[4] Box, G. E. P. and A. Luceño, Control by Monitoring and Feedback
Adjustment, John Wiley & Sons, New York, 1997.

[5] Brown, R. G., Smoothing, Forecasting and Prediction of Discrete Time
Series, Prentice-Hall, Englewood Cliffs, New Jersey, 1962.

[6] Capilla, C., A. Ferrer, R. Romero, and A. Hualda, “Integration
Statistical and Engineering Process Control in a Continuous
Polymerization Process”, Technometrics, 41, 1, 14-28, 1999.

[7] Harris, T. J. and W. H. Ross, “Statistical Process Control Procedures for
Correlated Observations”, The Canadian Journal of Chemical Engineering,
69, 48-56, 1991.

[8] Janakiram, M. and J. B. Keats, “Combining SPC and EPC in a Hybrid
Industry”, Journal of Quality Technology, 30, 3, 189-200, 1998.

[9] Janakiram, M., Statistical and Engineering Process Control integration
Strategies for Constrained Controllers, Thesis (Ph. D.) – Arizona State
University, 2001.

[10] MacGregor, J. F., “On-Line Statistical Process Control”, Chemical
Engineering Progress, 84, 21-23, 1988.

[11] MacGregor, J. F., “A Different View of the Funnel Experiment”, Journal
of Quality Technology, 22, 255-259, 1990.

[12] Mastrangelo, C. M. and E. C. Brown, “Shift Detection Properties of Moving
Centerline Control Chart Scheme”, Journal of Quality Technology, 32,
67-74, 2000.

[13] Mastrangelo, C. M. and D. C. Montgomery, “SPC with Correlated
Observations for Chemical and Process Industries”, Quality and
Reliability Engineering International , 11, 78-89, 1995.

[14] Matlab, A Language for Technical Computing, Computer Software, The
Math-Works, 1999.

[15] Messina, W. S., Strategies for the Integration of Statistical and
Engineering Process Control, Ph.D., Dissertation, Arizona State
University, Temp, AZ.,1992.

[16] Montgomery, D. C., Introduction to Statistical Quality Control, 4th
edition, John Wiley & Sons, New York, 1996.

[17] Montgomery, D. C., L. A. Johson, and J. S. Gardier, Forecasting and Time
Series Analysis, 2nd edition, Prentice-Hall, New York, 1990.

[18] Montgomery, D. C., J. B. Keats, G. C. Runger, and W. S. Messina,
“Integrating Statistical Process Control and Engineering Process
Control”, Journal of Quality Technology, 26, 79-87, 1994.

[19] Montgomery, D. C. and C. M. Mastrangelo, “Some Statistical Process
Control Methods for Autocorrelated Data”, Journal of Quality Technology,
23,179-193, 1991.

[20] Muth, J. F., ”Optimal Properties of Exponentially Weighted Forecasts”,
Journal of American Statistical Association, 55, 299-306, 1960.

[21] Nembhard, H. B. and M. S. Kao, “A Forecast-based Monitoring Methodology
for Process Transitions”, Quality and Reliability Engineering
International , 17, 307-321, 2001.

[22] Nembhard, H. B. and C. M. Mastrangelo, “Integrated Process Control for
Startup Operations”, Journal of Quality Technology, 30, 201-211, 1998.

[23] Ogata, K., Discrete-Time Systems, 2nd edition, Prentice-Hall, Englewood
Cliffs, New Jersey, 1994.

[24] Seborg, D. E., T. F. Edgar, and D. A. Mellichamp, Process Dynamics and
Control, 2nd edition, John Wiley & Sons, New York, 1989.

[25] Simulink, A Program for Simulating Dynamic Systems, Computer Software,
The Math-Works, Inc., Natick, MA., 1999.

[26] Tseng, S. and M. A. Banjamin, “Monitoring Autocorrelated Process
with an Exponentially Weighted Moving Average Forecast”, Journal of
Statistical Computation and Simulation, 50, 187-195, 1994.

[27] 韓曾晉, 適應控制系統, 科技圖書股份有限公司, 1992.