通常化學反應程序皆為非線性程序，各種以系統模式做為運算基礎所衍生出來的控制邏輯和策略，如模式預測控制(Model Predictive Control, MPC)，其整體控制表現與所採用的模式本身的預測準確度息息相關，對於一個高度非線性的化學反應系統，採用非線性模式，如人工智慧類神經網路，為基礎的控制方式顯然優於使用線性模式為控制基礎的策略。在人工智慧類神經網路十分流行的今天，其在建立經驗系統模式的應用上十分廣泛，簡單的架構和訓練方式是其優點，解決了推導真實系統數學模式所帶來的繁瑣與困難，並在大部分的情況下提供了一定程度的預測品質。人工智慧類神經網路近年來經過了各專門領域的努力，無論在技術上或應用上都有了相當大的改進和更深的認識，在化工程序控制方面，向前傳遞式和外層輸出回傳式類神經網路是建模時常被選用的兩種類型，兩者最主要的差別在於使用向前傳遞式類神經網路進行模式預測時，其輸入是採用線上量測設備所測得的數據，而外層輸出回傳式類神經網路則否，它改以本身輸出值回傳做為下一次輸入的方式取代，換言之，兩者最大的不同之處即在於輸入資料來源的取得，我們的研究發現，於每一個取樣控制的時刻不斷地自線上取得模式所需的輸入數據對模式預測控制而言具有相當負面的影響，尤其當我們要進行多步驟向前預測控測時同時夾雜著線上數據和模式預測數據的輸入項，其輸入資料的不一致性，在模式有誤差時將會破壞模式預測控制的品質，反觀完全使用模式預測輸出而不採用線上數據的外層輸出回傳式類神經網路則完全沒有這方面的問題。另外，不同於各式各樣的線性系統模式，一個非線性系統經驗模式準確度的優劣十分仰賴於當初用以建立模式的原始資料數據(或稱訓練學習資料)，這些數據資料本身的分布狀況和其所蘊涵的系統資訊量的多寡，對該非線性系統經驗模式的預測準確度而言扮演著十分重要的角色。
對人工智慧類神經網路模式而言，充足且分布均勻的訓練資料集合是十分重要的，因為其經驗模式的建立大多不使用任何先前知識，完全仰賴於訓練資料所提供的系統資訊，唯有資訊齊全的訓練資料庫才能建立起一個完善的人工智慧類神經網路模式，否則任何不屬於原始訓練學習樣本集合的新型態輸入資料數據將造成類神經網路模式產生若干不同程度且無法預知的外插運算，進而使得模式的預測值變得十分不可信賴，一般解決此類問題大多著力在提高模式本身的準確度，例如經由各種實驗設計方法和理論，以資料點均勻分布的觀念，來提供含有其它具有較多系統訊息的額外訓練資料，或是直接改良模式的訓練學習方法等等，然而並非所有由實驗設計法所建議的實驗都是可行的，尤其是在高度非線性系統中有很多區域的取樣十分困難，在現實的情況下有很多建議的實驗點都是很難去執行的，而且其所相對帶來的實驗成本更是一大問題，所以在真實的工業應用上仍難以獲得突破性的進展，事實上，就如同改進訓練學習方法一般，無論使用再怎樣精深的學理或技巧，在有限的訓練資料數據之下，面臨高度非線性系統時所能改進的空間亦將十分有限。

在論文中，我們首先針對向前傳遞式和外層輸出回傳式類神經網路二者間的差異與其對模式預測控制的影響予以分析和研究，在範例中實際展示其造成的影響，我們同時也提出了一個區域性知識指標，在實際程序控制中將之應用於分析動態人工智慧類神經網路模式中的每一個輸入項，藉由這些不屬於原始訓練學習樣本集合的新型態輸入所造成的指標值大幅滑落現象來提早辨識此類不適合利用模式預測控制的區域，再來我們為了進一步解決這些外插運算所造成的問題，確保控制系統的穩定，在不採用實驗設計額外增加實驗資料及變更學習邏輯的前提下，調整了控制架構並加入了一個專為穩定控制系統之用的穩定輔助控制器，採用雙控制器平行運算的架構，經由協調器□適當的權重設計來修正模式預測控制器的輸出。最後我們將此一方法應用在高度非線性的酸鹼中和控制程序中進測試和驗證，包括了程式模擬和前導工廠實驗兩部分，並獲得了相當優秀的控制成果。

1. Clarke, D. W., Mohtadi, C., and Tuffs, P. S., Generalized predictive control-Part 1: Basic algorithm. Automatica 23 (1987) 137-148.
2. Muske, K. R., and Rawlins, J. B., Model predictive control with linear models. A.I.Ch.E. Journal 39 (1993) 262-287.
3. te Braake, H. A. B., van Can, E. J. L., Scherpen, J. M. A., and Verbruggen, H. B., Control of nonlinear chemical processes using neural models and feedback linearization. Computers & Chemical Engineering 22 (1998) 269-276.
4. Wright, G. T., and Edgar, T. F., Nonliear model predictive control of a fixed-bed water-gas shift reactor: An experimental study. Computers & Chemical Engineering 18 (1994) 83-102.
5. Bodizs, A., Szeifert, F., and Chovan, T., Convolution model based predictive controller for a nonlinear process. Industrial Engineering and Chemistry Research 38 (1999) 154-161.
6. Noriega, J. R., and Wang, H., Adirect adaptive neural-network control for unknown nonlinear systems and its application. IEEE Transactions on Neural Network 1 (1998) 27-34.
7. Manner, B. R., Doyle, F. J., Ogunnaike, B. A., and Pearson, R. K., Non-linear model predictive control of simulated multivariable polymerization reactor using second-order volterra models. Automatica 32(9) (1996) 1285-1301.
8. Jang, S. S., Joseph, B., and Mukari, H., On-line optimisaion of constrained multivariable chemical processes. A.I.Ch.E. Journal 33 (1987) 26-41.
9. Patwardhan, A. A., Rawlings, J. B., and Edgar, T. F., Non-linear model predictive control. Chemical Engineering Communications 87 (1990) 123-135.
10. Qin, S. J., and Badgwell, T. A., An overview of industrial model predictive control technology. A.I.Ch.E. Symp. Ser. 93 (1997) 232.
11. Hussain, M.A., Review of the application of neural networks in chemical process control --- simulation and online implementation. Artificial Intelligence in Engineering 13 (1999) 55-68.
12. Dutta, P. and R.R. Rhinehart, Application of neural network control to distillation and an experimental comparison with other advanced controllers. ISA Transactions 38 (1999) 251-278.
13. Cutler, C.R. and R.L. Ramaker, Dynamic matrix control --- a computer control algorithm. AIChE 86th National Meeting, Houston, TX (1979).
14. Cutler, C.R. and R.L. Ramaker, Dynamic matrix control --- a computer control algorithm. Joint Automatic Control Conference, San Francisco, CA (1980).
15. Morari, M. and J.H. Lee, Model predictive control: past, present and future. Computers and Chemical Engineering 23 (1999) 667-682.
16. Garcia, C.E., D.M. Prett and M. Morari, Model predictive control: theory and practice --- a survey. Automatica 25 (1989) 335-348.
17. Maner, B.R., F.J. Doyle, B.A. Ogunnaike and R.K. Pearson, Non-linear model predictive control of a simulated multivariable polymerization reactor using second-order volterra model. Automatica 32 (1996) 1285-1301.
18. Sistu, P.B. and B.W. Bequette, Non-linear predictive control of uncertain processes: Application to a CSTR. A.I.Ch.E. Journal 37 (1991) 1711-1723.
19. MacMurray, J.C. and D.M. Himmelblau, Modeling and control of a packed distillation column using artificial neural networks. Computers and Chemical Engineering 19 (1995) 1077-1088.
20. Cybenko, G., Approximation by superpositions of a sigmoidal function. Mathematics of Control Signals and Systems 2 (1989) 303-314.
21. Gao, F.R., F.L. Wang and M.Z. Li, Predictive control for processes with input dynamic nonlinearity. Chemical Engineering Science 55 (2000) 4045-4052.
22. Bhat, N. and T.J. McAvoy, Use of neural nets for dynamic modeling and control of chemical process systems. Computers & Chemical Engineering 14 (1990) 573-582.
23. Nahas, E.P., M.A. Henson, and D.E. Seborg, Nonlinear internal model control strategy for neural network models. Computers & Chemical Engineering 16 (1992) 1039-1057.
24. Ljung, L. System Identification: Theory for the User. Prentice-Hall, Eaglewood Cliffs, NJ., (1987).
25. Economou, C.G..; Morari M.; Palsson, B.O. Internal model control. 5. Extension to nonlinear systems. Ind. Eng. Chem. Process Des. Dev. 25 (1986) 403-411.
26. Psichogios, D.C.; Ungar, L.H. Direct and indirect model based control using artificial neural networks. Ind. Eng. Chem. Res. 30 (1991) 2564-2573.
27. Lin, J.S. and S.S. Jang, Nonlinear dynamic artificial neural network modeling using an information theory based experimental design approach. Ind. Eng. Chem. Res.,37 (1998) 3640-3651.
28. Leonard, J.A., M.A. Kramer and L.H. Ungar, A neural network architecture that computes its own reliability. Computers & Chemical Engineering 16 (1992) 819-835.
29. Krishnapura, V.G., and A. Jutan, A neural adaptive controller. Chemical Engineering Science 55 (2000) 3803-3812.
30. Psaltis, D., A. Sideris and A.A. Yamamura, A multilayered neural network controller. IEEE Control Systems Magazine (1988) 17-21.
31. Nguyen, D.H. and B. Widrow, Neural networks for self-learning control systems. International journal of control 54(6) (1991) 1439-1451.
32. Loh, A.P., K.O. Looi and K.F. Fong, Neural network modeling and control strategies for a pH process, J. Proc. Control 5 (1995) 355-362.
33. , K.J. and T. , Automatic Tuning of PID Controllers, ISA, Research Triangle Park, NC, USA (1988).
34. Williams, G.L., R.R. Rhinehart, and J.B. Riggs, In-line process-model- based control of waste-water pH using dual base injection. Ind. Eng. Chem. Res. 29 (1990) 1254-1259.
35. Chan, H.C. and C.C. Yu, Autotuning of gain-scheduled pH control: an experimental study. Ind. Eng. Chem. Res. 34 (1995) 1718-1729.
36. Palancar, M.C., J.M. , J.A. and J.S. Torrecilla, Application of a model reference adaptive control system to the pH-control. Effects of lag and delay time. Ind. Eng. Chem. Res. 35 (1996) 4100-4110.
37. Palancar, M.C., J.M. , J.A. and J.S. Torrecilla, pH-control system based on artificial neural networks. Ind. Eng. Chem. Res., 37 (1998) 2729-2740.
38. Syu, M.J. and J.B. Chang, Recurrent backpropagation neural network adaptive control of penicillin acylase fermentation by arthrobacter viscosus. Ind. Eng. Chem. Res. 36 (1997) 3756-3761.
39. Frank, P.M. Entwurf von Regelhreisen mit Vorgeschirebenem Berhalten. G. Braun Verlag, Karlsruhe, 1974.
40. Morari, M., and Zafirou, E., Robust process control. (1989) Prentice-Hall.
41. Hunt, K. J., and Sbarbaro. D., Neural networks for nonlinear internal model control. IEE Proceedings-D 138(5) (1991) 431-438.
42. Narendra, K.S.; Parthasarathy, K. Identification and control of dynamical systems using neural networks. IEEE Trans. Neural Network 1 (1990) 4-27.
43. Haykin, S., Neural Networks: A Comprehensive Foundation. Prentice Hall International, Inc., 2nd edition. (1999).
44. Psaltis, D., A. Sideris, and A.A. Yamamura, A multilayered neural network controller. IEEE Control Systems Magazine 8 (1988) 17-21.
45. Yang, Y.Y. and D.A. Linkens, Adaptive neural-network-based approach for the control of continuously stirred tank reactor. IEE Proc.-Control Theory Appl. 141(5) (1994) 341-349
46. Bristol, E. On a new measure of interaction for multivariable process control. IEEE Trans. Auto. Control AC-11 (1966) 133-134.
47. Luyben,W.L. Process Modeling, Simulation, And Control for Chemical Engineers 2nd Ed., International Editions, McGraw-Hill, Inc., (1990).
48. Tsai, W.-Y., 2001. Artificial Neural Network Model Predictive Control on Packed Distillation Columns. Thesis of Master Degree, National Tsinghua University, Taiwan, 2001.
49. Wang, H., Y. Oh and E.S. Yoon, Strategies for modeling and control of nonlinear chemical processes using neural networks. Computers and Chemical Engineering 22 (1998) S823-S826.
50. Chien, I.L. and P.S. Fruehauf, Consider IMC tuning to improve controller performance. Chem. Engng Progr. 86 (1990) 33-41.
51. Specht, D.F., A general regression neural network. IEEE Transaction on Neural Networks 2 (1991) 568-576.