本論文使用Matlab/Simulink建置台北捷運系統模型，本模型將捷運系統分成三個部份：牽引系統、第三軌電壓和負載。牽引系統包含感應電動機與可變電壓可變頻控制、正弦脈波寬度調變和三相全橋變頻器。
負載部份包含了啟動電抗、運轉電抗和磁軌的梯度與旋度電抗。有四種操作模式用於速度調節，分別為加速、巡航、滑行、煞車。速度控制利用PID控制器建立於倒傳遞類神經網路中，利用類神精網路將PID控制器參數之調整自動化，並利用其線上學習的特性，使控制器能增強其適應性。
本論文中討論兩種情境，分別為梯度改變與負載改變，於情境中可得知控制器可穩定控制，當加速度與急動分別限制於 ±1 m/s2與 ±1 m/s3 以及急動限制於 ±0.8 m/s3 等情形。另外，本論文也示範再生儲能式煞車模式與滑行模式可成功地減少能源的耗損。

In this thesis, the actual train of Taipei Rapid Transit Systems (TRTS) is modeled with MATLAB/ Simulink. It consists of three main parts which are traction system, third rail voltage and the load. The traction system consists of induction motor with variable voltage variable frequency (VVVF) control, switching modes (SPWM, Quasi Six-step, and Sis Step), and three phase bridge inverter. The load consists of starting resistance, the running resistance during train movement, and gradient and curve resistances from the track profile. Four operation modes are used in the speed regulation which are acceleration, cruising, coasting, and braking.
Back Propagation Neural Network (BPNN) based PID is used as speed controller. This algorithm is the combination of Neural Network and PID. This technique makes PID tuning automatically since the tuning is done by the Neural Network. On-line learning capability of neural network can make the controller more adaptable.
Two case studies are presented to prove that the controller can work well regardless the gradient changes or load variations. The acceleration and jerk are limited within ± 1 m/s2 and ± 1 m/s3, respectively. Jerk limit ± 0.8 m/s3 which according to the TRTS regulation also implemented. Regenerative braking and coasting mode are successfully implemented to reduce the energy consumption.

[1] Network and Systems, http://english.trtc.com, TRTC, April 2014.
[2] Chen, C.S., Chuang, H.J., and Chen, J.L., “Analysis of Dynamic Load Behavior for Electrified Mass Rapid Transit Systems,” Industry Application Conference, Phoenix, pp. 992-998, Oct.1999.
[3] Tseng, K.H., and Shiao, Y.F., "The Analysis of Regenerative Braking Power for Taipei Rapit Transit Systems Electrical Multiple Unit," International Conference on Mechine Learning and Cybernetics, Xian, pp. 1947-1951, Jul.2012.
[4] Bing, G., Dong, H., and Zhang, Y., "Speed Adjustment Braking of Automatic Train Operation System based on Fuzzy-PID Switching Control," Sixth International Conference on Fuzzy Systems and Knowledge Discovery, Tianjin, pp. 577-580, Aug. 2009.
[5] Chen, X., Zhang, Y., and Huang, H., "Train Speed Control Algorithm Based on PID Controller and Single-Neuron PID Controller," Second WRI Global Congress on Intelligent System, Wuhan, pp. 107-110, Dec. 2010.
[6] Gu, Q., Tang, T., and Song, Y., "Neuroadaptive Variable Stucture Control of Mass Rapid Transit Trains," American Control Conference, San Francisco, pp. 775-779, Jun. 2011.
[7] Baier, M., Scholer, F., Zhao, D., Rulka, W., and Abel, D., "Sliding Mode Control for Comfort Improvement in Railway Vehicles,” 11th International Workshop on Variable structure System, Mexico City, pp. 130-135, Jun. 2010.
[8] Ke, B.R., Lin, C.L., and Lai, C.W., "Optimization of Train Speed Trajectory and Control of Mass Rapid Transit Systems," Control Engineering Practice, vol. 19, pp. 675-687, Apr. 2011.
[9] Su, S., Tang, T., Li, X., and Gao, Z., "Optimization of Multitrain Operations in a Subway System," IEEE Transaction on Intelligent Transportation Systems, vol. 15, no. 2, pp. 673-684, Apr. 2014.
[10] Yu, D., Lo, K.L.L., Wang, X., and Wang, X., "MRTS Traction Power Supply System Simulation Using Matlab/Simulink," Vehicular Technology Conference, pp. 308-312, 2002.
[11] Chang, C.S. and Xu, D.Y., "Differential Evolution Based Tuning of Fuzzy Automatic Train Operation for Mass Rapid Transit System," IEE Proceeding Electrical Power Application, vol. 147, no. 3, pp. 206-212, May 2000.
[12] Khmelnitsky, "On an Optimal Control Problem of Train Operation," IEEE Transactions on Automatic Control, vol. 45, no. 7, pp. 1257-1266, Jul. 2000.
[13] Lin, W.S. and Sheu, J.W., "Optimization of Train Regulation and Energy Usage of Metro Lines Using an Adaptive-Optimal-Control Algorithm," IEEE Transaction on Automation Science and Engineering, vol. 8, no. 4, pp. 855-864, Oct. 2011.
[14] Chiu, H.W., “Master Thesis: Simulation and Analysis of Train Performance Based on Actual MRT Operation Characteristics,” Taipei: NTUST, 2013.
[15] Shoujun, S. and Weiguo, L., "Application of Improved PID Controller in Motor Drive System," PID Control, Implementation and Tuning InTech, ch. 4, pp. 91-108, Apr. 2011.
[16] Gu, D. and Wang, G., "Application of PID Controller Based on BPNN in Temperature Control of Electrothermal Boiler," Advance in EECM: Springer, vol. 1, pp. 173-178, 2012.
[17] Zhang, J., Wang, S., Jiang Y., and Zhang X., "Induction Motor Control Based on BP Neural Network and Adaptive PID," 2nd International Conference on Computer Science and Network Technology, pp. 959-963, Dec. 2012.
[18] Yang, Y., Wu, Q., and Ding, S., "BP Neural Network based PID Controller for AC Servo Drive System," The 2nd International Conference on Intelligent Control and Information Processing, Harbin, pp. 1079-1081, Jul. 2011.
[19] Hsiao, Y.T., and Lin, K.C., "Measurement and Characterization of Harmonics on the Taipei MRT DC System," IEEE Transactions on Industry Application, vol. 40, no. 6, pp. 1700-1704, Nov. 2004.
[20] Bose, B.K., “Modern Power Electronics and AC Drives,” USA: Prentice Hall PTR, 2002.
[21] Mohan, N., Underland, T.M., and Robbins, W.P., “Power Electronics : Converters, Applications, and Design,” 3rd ed., USA: John Wiley & Sons, 2003.
[22] Versele, C., Deblecker, O., and Lobry, J., "Implementation of Induction Motor Drive Control Schemes in MATLAB/ Simulink/ dSPACE Environment for Educational Purpose," MATLAB for Engineers - Application in Control, Electrical Engineering, IT and Robotics. www.intechopen.com, pp. 365-386, Oct. 2011.
[23] Chuang, H.J., Chen, C.S., Fan, L.J., and Ho, C.Y., "Enhancement of Power System Operation for Taipei MRT Network," International Conference on Power System Technology, Chongqing, pp. 1-7, Oct. 2006.
[24] Patel, H.S., and Hoft, R.G., "Generalized Techniques of Harmonic Elimination and Voltage Control in Thryristor Inverters: Part I-Harmonic Elimination," IEEE Transactions on Industry Applications, vol. 1A-9, no. 3, pp. 310-317, May 1973.
[25] Patel,H.S. and Hoft,R.G., "Generalized Techniques of Harmonic Elimination and Voltage Control in Thyristor Inverters: Part II-Voltage Control Techniques," IEEE Transactions on Industry Applications, vol. 1A-10, no. 5, pp. 666-673, Sep. 1974.
[26] Sahali, Y. and Fellah, M.K., "Selective Harmonic Eleiminated Pulse Width Modulation Technique(SHE PWM) applied to the Three-lecel Inverter/ Converter," International Symposiom on Industrial Electronics, vol. 2, pp. 1112 – 1117, Jun.2003.
[27] Tuncer,S., Tartar, Y. and Guldemir, H., "A SHEPWM Technique with Constant v/f for Multilevel Inverter," Journal of Polytechnic, vol. 8, no. 2, pp. 123-130, 2005.
[28] Yadav, A., “Traction Choices: Overhead AC vs. Third Rail DC,” IRJ:International Railway Journal. [Online], http://www.railjournal.com/index.php/rolling-stock/traction-choices-overhead-ac-vs-third-rail-dc.html, Feb. 2013.
[29] Su, I.,“Vehicles of MRT and THSR,” Citycat's Railway [Online], http://citycat.theweb.tw/tmrt.htm, Jan. 2001.
[30] Hay,W.W., “Railroad Engineering,” 2nd ed., John Willey & Sons, 1982.
[31] Shen, Z.B., "MRT and LRT Supply System Computerized Simulation(Upper)-Train Running Simulation," MRT Technology Taiwan North Transit Authority, pp. 161-184, Aug. 1991.
[32] Rochard, B.P. and Schmid, F., "A Review of Methods to Measure and Calculate Train Resistance," Journal of Rail and Rapid Transit, pp. 185-199, Jul. 2000.
[33] Railway Technical Web Pages. [Online], http://www.railway-technical.com/whlbog.shtml, Jun. 2014.
[34] Ho, T. K. and Wong, K.K., "Coast Control for Mass Rapid Transit Railways with Searching Methods," IEE Proceeding Electrical Power Applications, vol. 151, no. 3, pp. 365-376, May 2004.
[35] Chang,C.S. and Sim, S.S., "Optimising Train Movements Through Coast Control Using Genetic Algorithms," IEE Proceeding Electrical Power Applications, vol. 144, no. 1, pp. 65-73, Jan. 1997.
[36] Chuang, H. J., Chen, C.S., Lin, C.H., and Hsieh, C.H., "Design of Optimal Coasting Speed for MRT Systems Using ANN Models," IEEE TRANSACTION ON INDUSTRY APPLICATIONS, vol. 45, no. 6, pp. 2090-2097, Nov. 2009.
[37] J. F. Chen, R. L. Lin, and Y. C. Liu, "Optimization of an MRT Train Schedule: Reducing Maximum Traction Power by Genetic Algorithm," IEEE TRANSACTIONS ON POWER SYSTEMS, vol. 20, no. 3, pp. 1366-1372, Aug. 2005.
[38] Bansal, H.O., Sharma, R., and Shreeraman, P.R., "PID Controller Tuning Techniques: A review," Journal of Control Engineering and Technology(JCET), vol. 2, no. 4, pp. 168-176, Oct. 2012.
[39] Araki, M., "PID Control in Control System, Robotics, and Automation," Encyclopedia of Life Suppoet System(EOLSS), UK, Oxford: Eolss, vol. 2, 1989.
[40] Karray, Fakhreddine, and Silva, D., “Soft Computing in Intelligent Systems,“ Harlow, England: Pearson Education, 2004.
[41] Balestrino, A., Caiti, A., Calabro, V., Crisostomi, E., and Landi, A., "From Basic to Advanced PI Controller," Advances in PID Control, www.intechopen.com, ch. 5, pp. 85-100, 2011.
[43] Leverington, D., ”A Basic Introduction to Feedforward Backpropagation Neural Networks,” [Online], http://www.webpages.ttu.edu/dleverin/neural_network/neural_networks.html, May 2014.
[44] Kusumadewi, S., “Artificial Intelligent,” Yogyakarta, Indonesia: Graha Ilmu, 2003.
[45] Maghfiroh, H., Wahyunggoro, O., Cahyadi, A.I., and Praptodiyono, S., "PID Hybrid Tuning to Improve Control Performance in Speed Control of DC Motor Base on PLC," 3rd International Conference on Instrumentations Control and Automation, Bali, , pp. 233-238, Aug. 2013.
[46] Ke, B. R., Lin, C.L., and Yang, C.C., "Optimization of Train Energy-Efficient Operation for Mass Rapid Transit Systems," IET Intelligent Transport Systems, vol. 6, no. 1, pp. 58-66, 2012.