因應磁浮主軸系統動態具開路不穩定的特性，本論文發展出一套閉迴路鑑別方法，藉此獲得精確的系統模型作為控制器設計的依據。此鑑別方法是基於分散式解耦的控制架構，將多輸入多輸出的系統動態解耦為兩個單輸入單輸出系統來簡化鑑別程序。鑑別過程採用偽隨機二進位序列作為激發訊號，透過基於輸出誤差模型的參數估測演算法來得到參數化模型。為確保演算過程中模型預測誤差能收斂，估測演算法中需引入一濾波器，而本論文特別針對參數估測的濾波器提出一項系統化的設計程序。此程序是基於線性矩陣不等式的求解，利用數值計算保證引入濾波器後的演算法能穩定且收斂。論文提出的控制與鑑別方法先利用模擬驗證其可行性，再將其實際施行於五軸磁浮主軸平台上，並以實驗頻率響應來對鑑別所得之參數化模型進行驗證。

In response to the open-loop instability of magnetic bearing systems, this thesis develops a closed-loop identification method to obtain an accurate system model for controller design purposes. The identification method is based on a decentralized and decoupling control architecture, which allows one to decouple the two-input-two-output system into two single-input single-output systems so as to simplify the identification process. The identification process uses a pseudo-random binary sequence as an excitation signal. A parameterized model is obtained through a parameter estimation algorithm based on the output error model. To ensure that the model prediction error can converge during the computation process, a filter is incorporated into the estimation algorithm. The filter can be synthesized systematically by the solving a set of linear matrix inequalities. The performance of the proposed closed-loop identification method is firstly verified by simulations. Then it is implemented on a five-axis magnetic bearing platform and the parameterized model obtained from the identification are compared to the experimental frequency response.

[1] Yeh, T. J. (1996). Modeling, analysis and control of magnetically levitated rotating machines (Doctoral dissertation, Massachusetts Institute of Technology).
[2] Schweitzer, G., & Maslen, E. H. (2009). Magnetic bearings: theory, design, and application to rotating machinery (Vol. 2009). Berlin: Springer.
[3] Noshadi, A., Shi, J., Lee, W. S., Shi, P., & Kalam, A. (2016). System identification and robust control of multi-input multi-output active magnetic bearing systems. IEEE Transactions on control Systems technology, 24(4), 1227-1239.
[4] Khader, S. A. (2015). System Identification of Active Magnetic Bearing for Commissioning.
[5] Shi, J., & Revell, J. (2002, October). System identification and re-engineering controllers for a magnetic bearing system. In 2002 IEEE Region 10 Conference on Computers, Communications, Control and Power Engineering. TENCOM'02. Proceedings. (Vol. 3, pp. 1591-1594). IEEE.
[6] Hu, L., & Ming, D. (2019, March). Research on Vibration of Magnetic Suspension Rotor System Caused By Magnetic Bearing Model Error—Closed-Loop Paramter Identification Method. In IOP Conference Series: Materials Science and Engineering (Vol. 493, No. 1, p. 012012). IOP Publishing.
[7] Medina, J., Parada, M., & Medina, L. (2004, June). A neural network-based closed loop identification of a magnetic bearings system. In ASME Turbo Expo 2004: Power for Land, Sea, and Air (pp. 593-598). American Society of Mechanical Engineers Digital Collection.
[8] Kothari, K., Mehta, U., & Singh, N. (2018, November). Practical test for closed-loop identification and control on magnetic levitation system: A fractional-order approach. In 2018 15th International Conference on Control, Automation, Robotics and Vision (ICARCV) (pp. 1805-1810). IEEE.
[9] Landau, I. D., & Boumaiza, K. (1996). An output error recursive algorithm for unbiased identification in closed loop. IFAC Proceedings Volumes, 29(1), 4144-4149.
[10] Landau, I. D., & Zito, G. (2007). Digital control systems: design, identification and implementation. Springer Science & Business Media.
[11] Landau, I. D., Lozano, R., M'Saad, M., & Karimi, A. (2011). Adaptive control: algorithms, analysis and applications. Springer Science & Business Media.