永磁同步馬達因具備高效率及高功率密度的特點，加上磁性材料與電力電子技術的進步，已逐漸取代傳統感應馬達。然而，當電機於高速運轉時，所引發的損失會促使轉子溫度上升，在散熱能力有限導致熱量不斷累積的情況下，磁鐵本身因天生材料的特性會產生不可逆的退磁現象，一般此現象並非瞬間發生，而是長時間不斷劣化的過程。在工業產線使用的馬達，當馬達發生故障，就有可能造成產線停擺及巨大損失，本論文探討退磁狀態監測技術，從磁鐵材料特性及馬達損失分析，利用控制技術可估測出馬達的磁交鏈，所得之虛擬量測的訊號以及可直接量測之訊號，經由長短期記憶神經網路(Long Short-Term Memory, LSTM)運算後端監測分析平台。本研究之實驗是以機房風扇馬達作為應用情境，並使用硬體在環系統取代真實馬達，可即時調整參數改變馬達健康狀態，透過微處理器控制此虛擬馬達系統，利用訊號擷取取得物理量，上傳至國網中心分析平台，進行遠端監測及健康狀態預估，本研究藉由實驗結果驗證所提出之估測器及健康狀態預估演算法則之可行性。

When PMSM operates under high-speed condition, the loss of rotor increases because of time harmonics and spatial harmonics. Thus, more heat will be produced and accumulated. However, the heat dissipation of motor is limited. Excessive heat loss makes motor temperature rise dramatically, which may lead to permanent magnet irreversible demagnetization and decrease the motor performance. In order to find out the reason of demagnetization, this paper analyzes material properties of PM, and derive the relationship between temperature and flux linkage. The fault does not occur instantaneously but in a gradually degrading process. Therefore, a technique to monitor demagnetization condition of PMSM is proposed. It consists of a magnetic flux observer and a long short-term memory (LSTM) algorithm. The observer is regarded as indirect measurement. Take the information obtained from direct and indirect measurement as inputs of LSTM, and diagnosis the health condition of PM. An experimental platform composed of Hardware in the Loop (HIL), microcontroller and data acquisition (DAQ) devices is established to verify the proposed flux observer and monitoring technique.

[1]H. Kagermann, W. D. Lukas, and W. Wahlster, "Industrie 4.0: mit dem internet der dinge auf dem weg zur 4. industriellen revolution," VDI nachrichten, vol. 13, no. 1, 2011.
[2]K. Imamura, M. Sanada, S. Morimoto, and Y. Inoue, "Improvement of demagnetization by rotor structure of IPMSM with dy-free rare-earth magnet," International Conference on Electrical Machines and Systems (ICEMS), pp. 1-6, 2012.
[3]M. Satoh, S. Kaneko, M. Tomita, and S. Doki, "Study to reduce the amount of magnetic materials in IPM motor which focuses on the rare earth magnet usage and demagnetization," International Conference on Electrical Machines and Systems (ICEMS), pp. 195-199, 2015.
[4]J. R. R. Ruiz, J. A. Rosero, A. G. Espinosa, and L. Romeral, "Detection of demagnetization faults in permanent-magnet synchronous motors under nonstationary conditions," IEEE Transactions on Magnetics, vol. 45, no. 7, pp. 2961-2969, 2009.
[5]A. G. Espinosa, J. A. Rosero, J. Cusido, L. Romeral, and J. A. Ortega, "Fault detection by means of Hilbert–Huang transform of the stator current in a PMSM with demagnetization," IEEE Transactions on Energy Conversion, vol. 25, no. 2, pp. 312-318, 2010.
[6]M. Zhu, W. Hu, and N. C. Kar, "Acoustic noise-based uniform permanent-magnet demagnetization detection in SPMSM for high-performance PMSM drive," IEEE Transactions on Transportation Electrification, vol. 4, no. 1, pp. 303-313, 2017.
[7]I. H. Kao, W. J. Wang, Y. H. Lai, and J. W. Perng, "Analysis of permanent magnet synchronous motor fault diagnosis based on learning," IEEE Transactions on Instrumentation and Measurement, no. 99, pp. 1-15, 2018.
[8]S. Moosavi, A. Djerdir, Y. A. Amirat, and D. Khaburi, "Demagnetization fault diagnosis in permanent magnet synchronous motors: a review of the state-of-the-art," Journal of magnetism and magnetic materials, vol. 391, pp. 203-212, 2015.
[9]I. K&J Magnetics. [Online]. Available: https://www.kjmagnetics.com/bhcurves.asp.
[10]J. X. Shen, H. Hao, M. J. Jin, and C. Yuan, "Reduction of rotor eddy current loss in high speed PM brushless machines by grooving retaining sleeve," IEEE Transactions on Magnetics, vol. 49, no. 7, pp. 3973-3976, 2013.
[11]劉昌煥，「交流電機控制：向量控制與直接轉矩控制原理」，臺灣東華，2006。
[12]劉子瑜，「基於弦波電流驅動於永磁同步馬達電流迴路控制之研究」， 碩士論文，電機工程學系碩博士班，國立成功大學，台南市，2009。
[13]G. Ellis, "Observers in control systems: a practical guide," Elsevier, 2002.
[14]林睿晟，「利用反電動勢觀測器調校馬達參數」，馬達電子報，第828期，2019。
[15]Z. C. Lipton, D. C. Kale, C. Elkan, and R. Wetzel, "Learning to diagnose with LSTM recurrent neural networks," arXiv, 2015.
[16]M. Yuan, Y. Wu, and L. Lin, "Fault diagnosis and remaining useful life estimation of aero engine using LSTM neural network," International Conference on Aircraft Utility Systems, pp. 135-140, 2016.
[17]S. Hochreiter and J. Schmidhuber, "Long short-term memory," Neural computation, vol. 9, no. 8, pp. 1735-1780, 1997.
[18]凱登智動科技，「MR series 說明書」[Online]. Available: https://www.gathertech.net/.
[19]Y. C. Lin, "Development of advanced manufacturing cloud of things (AMCoT)—a smart manufacturing platform," IEEE Robotics and Automation Letters, vol. 2, no. 3, pp. 1809-1816, 2017.
[20]R. Steinbrecher and R. Schmidt, "Data center environments," ASHRAE Journal, vol. 53, pp. 42-49, 2011.