臺灣博碩士論文加值系統

一般控制理論的分析與應用，都需要將系統的數學描述模型，根據模型輸出/輸入資料或物理特性先求出，但往往在建立描述系統的模型數學過程中，因為模型本身所具有的複雜性、不確定性(Uncertainty)等等的因素，而無法得到精確的系統模型，使得設計出的系統性能無法達到預期的效果與目標。

本論文之目地在於利用，類神經網路倒傳遞演算法所擁有簡單的架構和訓練方式等的優點，來完成模型辨識的工作，解決推導真實系統數學模式所帶來的繁瑣與困難，並將此辨識完成，具備有模型特徵的神經網路，應用在模式參考適應性控制器上使得受控體輸出響應能完全跟隨參考模式的軌跡，將神經網路在控制方面的應用作初步的探討。

For the analysis and application of control-theory, It is necessary to
model the system to a mathematic platform, according to the physical, input and output characteristic of the system.

Unfortunately, the complexity of the system and uncertainty of environment cause that the model cannot exactly and accurately performs all the behavior of the target system. Therefore, the implementation cannot achieve our expected performance due to inaccurate modeling.

This paper will apply the advantages of simple architecture and operation
of backpropagation to setup the system. It avoids the complexity of the mathematic modeling. the implementation of nervous-network model, whose response completely follows up the reference model, is applied to the model reference adaptive controller.

第一章 序論．．．．．．．．．．．．．．．．．．．．．．-01-
1.1 研究的背景與動機．．．．．．．．．．．．．．．．-01-
1.2 相關文獻回顧．．．．．．．．．．．．．．．．．．-02-
1.3 研究內容．．．．．．．．．．．．．．．．．．．．-03-

第二章 神經網路理論與架構．．．．．．．．．．．．．．．-05-
2.1基本神經網路介紹．．．．．．．．．．．．．．．．-05-
2.2神經網路架構．．．．．．．．．．．．．．．．．．-08-
2.2.1神經元模型組成單元．．．．．．．．．．．．．．-09- 2.3神經網訓練過程．．．．．．．．．．．．．．．．．-11-
2.4倒傳遞演算法(Backpropagation) ．．．．．．．．．-11-
2.6倒傳遞演算法的改善．．．．．．．．．．．．．．．-15-
2.6.1動量(Momentum)倒傳遞演算法．．．．．．．．-19-
2.6.2可變學習速率倒傳遞演算法．．．．．．．．．-20-
2.6.3 彈性(Resilient)的倒傳遞演算法．．．．．．-21-
2.6.4共軛梯度(Conjugate Gradient)演算法．．．．-22-
2.6.5擬牛頓(Quasi-Newton)演算法．．．．．．．．-24-
2.6.6Levenberg-Marquardt 演算法．．．．．．．．-25-

第三章 系統辨識與模型參考適應性控制．．．．．．．．．．-26-
3.1系統辨識基本概念．．．．．．．．．．．．．．．．-27-
3.2系統辨識方法．．．．．．．．．．．．．．．．．．-29-
3.3模型參考適應性控制架構．．．．．．．．．．．．．-30-

第四章 系統辨識與模式參考輸出預測模擬結果．．．．．．．-33-
4.1前置處理．．．．．．．．．．．．．．．．．．．．-33-
4.2系統辨識．．．．．．．．．．．．．．．．．．．．-35-
4.2.1倒傳遞演算法訓練神經網路．．．．．．．．．．-35-
4.2.2動量&可變學習速率演算法訓練神經網路．．．．-39-
4.2.3彈性(Resilient)倒傳遞演算法訓練神經網路．．-42-
4.2.4 神經網路訓練效果．．．．．．．．．．．．．-46-
4.3模式參考控制器訓練．．．．．．．．．．．．．．．-48-

第五章結論．．．．．．．．．．．．．．．．．．．．．．．-50-
5.1結論．．．．．．．．．．．．．．．．．．．．．．-50-

[1]W. McCulloch and W. Pitts,” A logical calculus of the ideas immanent in nervous activity,” Bull. Math. Biophy. ,vol.5 pp.115.133,1943

[2]F. Rosenblatt,” The Perceptron: A Probabilistic model for information storage and organization in the brain," Psychological Review,vol.65 pp.386-408,1958

[3]B. Widrow, M.E.Hoff, “Adaptive switching circuit”1960 IRE WESCON Convention Record, New York: IRE Park,4,pp.96 104,1960

[4]T. Kohonen, “Self-organized formation of topologically correct feature maps,” Biol. Cybern., no. 43, pp. 59-69,1982.

[5]G. A. Carpenter, S. Grossberg, and D. B. Rosen, “ART 2-A:An adaptive resonance algorithm for rapid category learning and recognition,” Neural Net., vol. 4, pp. 493-504, 1991.

[6]J. J. Hopfield, “Unlearning has a stabilizing effect in collective memories,” Nature., vol. 304, pp. 158, 1983.

[7]D. E. Rumelhart, G. E. Hinton and R. J. Williams, “Learning internal representations by error propagation,” PARALLEL distributed Processing:Explorations in the Microstructures of Cognition, Vol. 1:Foundations. D. E. Rumelhart and J. L.McCleland, Eds. Cambridge, MA, MIT Press,1969.

[8]D.E. Rumelhart and J.L. McClelland, eds. Parallel Distributed Processing: Explorations in the Microstructure of Cognition,vol.1,Cambridge,MA:MIT Press,1986

[9]R.A. Jacobs,”Increased rates of convergence through learning rate adaptation”Neural Networks,vol.1 pp295-308 1988

[10]羅華強, “ 類神經網路 : MATLAB 的應用” , 鈦思科技, 2001.

[11]史科忠 ,“ 神經網路控制理論 ” 西安交通大學出版社 , 1997

[12]陳燕慶 ,”神經網路理論及其在控制工程中的應用 ”儒林 1992 [民81]

[13]Hagan Demuth Beale, “類神經網路設計” 審校:汪惠,健普林斯頓 ,2004

[14]Narendra, K. S., and Parthasarathy, K. “Identification and control of dynamical systems using neural networks,” IEEE Transactions on Neural Networks, Vol. 1, No. 1, pp. 4-27, March 1990.

[15]Kuntanapreeda, S.; Gundersen, R.W.; Fullmer, R.R.; “Neural network model reference control of nonlinear systems”. Vol 1.2, Page(s):94 - 99 June 1992

[16]Mazumdar, S.K.; Lim, C.C.”Adaptive controller for marginally stable nonlinear systems using neural networks” IEEE vol.1 Page(s):535 - 539 Nov. 1992

[17]Liang Jin; Nikiforuk, P.N.; Gupta, M.M.; ”Fast neural learning and control of discrete-time nonlinear systems”,IEEE Vol 25, Page(s):478 - 488 March 1995

[[18]]C.Charalambous,”Conjugate gradient algorithm for efficient training of artificial neural network” IEE Proceedings, vol.139, no.3,pp.301-310.