本文主旨以現場可規劃邏輯陣列(FPGA) 作為發展嵌入式控制系統主要的硬體核心; 以壓電致動器系統為例, 探討進階數位控制器實現的諸多議題。其內容包含了非以系統數學模型設計(Non-Model-based design)PID 控制器與類神經網路(Neural Network)控制器, 以及系統數學模型為基礎設計(Model-based design)之反覆控制器(Repetitive Control) 。有關反覆控制器設計所需的數學模型,
本文在10 kHz及100 kHz的取樣頻率下, 以最小平方模型(Auto eRegressive eXogeneous)進行系統鑑別以供控制器設計。此外本研究也針對壓電致動器的磁滯現象加以建模並進行補償, 以提升精密控制定位效能。在嵌入式系統FPGA 晶片的數位控制器實現方面, 硬體採用美商國家儀器(National Instruments)所生產CompactRIO(cRIO) 嵌入式硬體, 其內部由Real-Time Module 、FPGA Module
以及DAQ Module 所構成; 相關韌體撰寫則以LabVIEW 軟體作為溝通與控制器開發介面。其開發內容包括數值運算分析、程式優化及演算函式轉為電路型式實現方法等; 除此之外, 本文並探討FPGA 晶片內部邏輯電路實現高等數位控制演算法需探討之特性; 例如定點數運算、有限精度運算以及溢位等。藉由上述分析與討論, 本研究成功地實現類神經網路與反覆控制於所選用之嵌入式控制系統, 並獲得比傳統PID 控制器更好的壓電致動器追跡控制效能。

This thesis aims to develop necessary techniques in an embedded control system by using the FPGA based platform. The study takes piezoelectric actuator as a control system example and investigates several issues for digital controller implementation. The applied controls include non model based design such as PID control and neural network control. Moreover, repetitive control, a model based design approach, is also implemented on the piezo actuated system for periodic signal tracking at 10 kHz and 100 kHz sampling rate, respectively. The mathematical model for repetitive controller design is obtained by using Auto-Regressive eXogenous (ARX) algorithm, a system identification technique, for the repetitive controller design. To compensate for the hysteresis effect in the piezo actuated system, an inverse hysteretic compensator is also applied to further improve the tracking performance. The digital control algorithms were realized using embedded hardware CompactRIO developed by National Instruments. The controller includes Real Time module, FPGA module, and DAQ module. We specifically develop the controller firmware under an FPGA environment using LabVIEW software language and obtain extreme fast sampling rate and better control performance than similar works in the literature. Because the FPGA controller is a fixed point processor, design issues such as finite word length precision, overflow, and program optimization should be carefully considered. After performing fixed point analysis, this research successfully implements the advanced controllers on the selected embedded control system. The experimental results also demonstrate that the applied neural network control and repetitive control achieve better control performance than traditional PID control.

[1] Callender, A. and Hartree, DR and Porter, A. Time-lag in a control sys-
tem. Philosophical Transactions of the Royal Society of London. Series A,
Mathematical and Physical Sciences, pp. 415-444, 1936.
[2] Ziegler, JG and Nichols, NB. Optimum settings for automatic con-
trollers. Transactions of the ASME, 1942.
[3] Rosenblatt, F. The perceptron: A probabilistic model for information storage
and organization in the brain. Psychological review, vol. 65, no. 6, pp. 386-
408, 1958.
[4] Rumelhart, DE and Hinton, GE and Williams, RJ. Learning internal repre-
sentation by error propagation, Parallel Distributed Processing, Vol. 1: foun-
dations, DE Rumelhart and JL McClelland, eds. MIT Press, Cambridge,
Mass, 1986.
[5] McClelland, J.L. and Rumelhart, D.E. and others. Parallel distributed pro-
cessing: Explorations in the microstructure of cognition. MIT press Cam-
bridge, MA, 1986.
[6] Psaltis, D. and Sideris, A. and Yamamura, AA. A multilayered neural net-
work controller. IEEE Control Systems Magazine, vol. 8, no. 2, pp. 17-
21, 1988.
[7] Xu, Q. and Li, Y. Radial basis function neural network control of an
XY micropositioning stage without exact dynamic model. IEEE/ASME In-
ternational Conference on Advanced Intelligent Mechatronics, pp. 498-
503, 2009.
[8] Meireles, M.R.G. and Almeida, P.E.M. and Simoes, M.G. A comprehen-
sive review for industrial applicability of artificial neural networks. in IEEE
transactions on Industrial Electronics, vol. 50, no. 3, pp. 585-601, 2003.
[9] Tomizuka, M. and Tsao, T.C. and Chew, K.K. Analysis and synthesis of discrete-time repetitive controllers. Journal of Dynamic Systems, Measure-
ment and Control, vol. 111, no. 3, pp. 353-358, 1989.
[10] Tomizuka, M. Zero phase error tracking algorithm for digital control. Jour-
nal of Dynamic Systems, Measurement and Control, vol. 109, no. 1, pp. 65-
68, 1987.
[11] 姚武松. 高速工具機之線性伺服系統設計. 國立成功大學機械工程學系博士
論文, 2002.
[12] Bakhri, S. and Ertugrul, N. and Soong, WL and Al-Sarawi, S. Investigation
and development of a real-time on-site condition monitoring system for in-
duction motors. Universities Power Engineering Conference, pp. 1-6, 2007.
[13] Leong, PHW and Tsang, PK and Lee, TK. AFPGA based Forth micropro-cessor. Published by the IEEE Computer Society, pp. 254, 1998.
[14] Kung, Y.S. and Fung, R.F. and Tai, T.Y. Realization of a Motion Control IC
for X−Y Table Based on Novel FPGA Technology. IEEE Transactions on
Industrial Electronics, vol. 56, no. 1, pp. 43-53, 2009.
[15] Monmasson, E. and Cirstea, MN. FPGA design methodology for indus-
trial control systems-A review. IEEE Transactions on industrial Electron-
ics, vol. 54, no. 4, pp. 1824-1842, 2007.
[16] 林容益編著. FPGA 數位IC 電路設計應用及實驗. 國立中興大學, 2007.
[17] 張曉倩. 以行為模型建立三角積分數位類比轉換器之準確時間響應的研
究. 國立中央大學電機工程研究所碩士論文, 2009.
[18] Hsu, S.P. Linear Induction Motor Drive System Based on Backstepping
Controller. Electrical Engineering, 2000.
[19] 孫益君. 以PDA 為平台之語音辨識應用系統開發. 私立中原大學資訊工程學
系碩士論文, 2004.
[20] Chang, C.Y. and Fu, S.Y. Image classification using a module RBF neural
network. IEEE Computer Society, 2006.
[21] 張斐章, 張麗秋. 類神經網路. 東華書局, 2005.
[22] Chen, S. and Mulgrew, B. and Grant, PM. A clustering technique for
digital communications channelequalization using radial basis function net-
works. IEEE Transactions on Neural Networks, vol. 4, no. 4, pp. 570-
590, 1993.
[23] Chen, P. and Tsai, H. and Lin, C.J. and Lee, C.Y. FPGA Realization of
a Radial Basis Function Based Nonlinear Channel Equalizer. Advances in
Neural Networks, pp. 320-325, 2005.
[24] Kuo, S.M. and Lee, B.H. and Tian, W. Real-time digital signal processing:
implementations and applications. John Wiley & Sons, 2006.
[25] Kuhnen, K. and Janocha, H. Inverse feedforward controller for complex
hysteretic nonlinearities in smart-material systems. Control and Intelligent
Systems, vol. 29, no. 3, pp. 74-83, 2001.
[26] CEDRAT TECHNOLOGIES. Piezo Actuators, Drivers & Controllers Cat-
alogue 2007. CEDRAT TECHNOLOGIES, 2007.
[27] Huang, P.Y. System Identification and Control of a Paper Machine. National
Pingtung University of Science and Technology Master Thesis, 2007.
[28] Dorf, R.C and Bishop, R.H. Modern control systems. Prentice Hall, 2008.