本文旨在研究度量學習輔助之域適應網路(Metric-Learning-Assisted Domain Adaptation, MLA-DA)於感應馬達跨域故障診斷性能，將其與其餘類型之轉移學習模型進行比較，並研究其餘不同情境下的應用，包含同一馬達間跨負載狀態、不同機組容量馬達之跨容量，及不同負載狀態及機組容量之綜合跨負載與容量的跨域故障診斷。透過案例設計及結果分析，驗證度量學習輔助之域適應模型於馬達跨域故障診斷應用的可行性。
基於相關文獻，本文選擇深度域混淆(Deep Domain Confusion, DDC)、域對抗神經網路(Deep Adversarial Neural Network, DANN)、深度子域適應(Deep Subdomain Adaptation Network, DSAN)及度量學習輔助之域適應等不同類型的轉移學習模型進行比較。此外，本研究以兩馬力及五馬力感應馬達作為故障診斷對象，研製定子匝間短路、轉子斷條、軸承鑽孔及不對心故障四種瑕疵馬達模型，實現三相感應馬達跨域故障診斷。再者，本研究根據準確度結果固定訓練數據之感測器軸向及圖像中資料筆數，並以在該設定下，生成之二維時域圖像作為輸入數據，對度量學習輔助之域適應及其餘三種模型進行訓練，並比較四種模型於跨容量、跨負載及綜合跨容量與負載故障診斷成效，理解不同模型於不同故障類型的表現差異，並根據結果確定各模型於馬達跨域故障診斷應用的優勢。
最後，根據實驗結果顯示，度量學習輔助之域適應模型於跨負載故障診斷平均準確度可達98.93%，跨容量故障診斷平均準確度可達91.38%，綜合跨負載及容量案例中，準確度最高可達88.7%，且以F1-score來看，度量學習輔助之域適應模型對於某幾種故障型態性能優於深度子域適應模型。在上述三種實驗結果中，度量學習輔助之域適應模型之分類準確度皆高於其餘三種模型，如此便完成度量學習輔助之域適應模型於感應馬達跨域故障診斷之研究。

This paper aims to investigate the performance of Metric-Learning-Assisted Domain Adaptation (MLA-DA) in cross-domain fault diagnosis of induction motors. It was compared with other types of transfer learning model, and its application in different scenarios, including cross-loading condition within the same motor, cross-capacity among motors in different power ratings, and cross-domain fault diagnosis involving both loading condition and power ratings, were examined. Through result analysis, the feasibility applying the MLA-DA model in cross-domain fault diagnosis was validated.
First, relevant literature was reviewed to understand the background of cross-domain fault diagnosis in induction motors, the classification of transfer learning models, and common types of motor fault. Based on it, a comparison was made among different types of models, including Deep Domain Confusion (DDC), Domain Adversarial Neural Network (DANN), Deep Subdomain Adaptation Network(DSAN), and MLA-DA. Additionally, this study focused on two horsepower and five horsepower induction motors as the subjects of fault diagnosis, and developed four types of defective motor models, including stator short circuit, rotor broken bar, drilled bearing, and misalignment, to facilitate cross-domain fault diagnosis of three-phase induction motors.
Furthermore, based on the accuracy results, the sensor axial direction and the number of data samples in the training data were fixed, and two-dimensional time-series images generated under this configuration were used as input data to train MLA-DA and other three models. The performance of the four models in cross-capacity, cross-load, and integrated cross-capacity and load fault diagnosis was compared to understand the performance differences among different models for different fault types. The advantages of each model in the application of cross-domain fault diagnosis for motors were determined based on the results.
Finally, according to the experimental results, the MLA-DA model can reach an average accuracy of 98.93% for cross-load fault diagnosis, 91.38% for cross-capacity fault diagnosis, and 88.7% for the integrated cross-load and capacity cases, and the MLA-DA model outperforms the DSAN model in terms of the F1-score for certain types of motor faults. In the above three experimental results, the classification accuracy of the MLA-DA model is higher than the other three models, which completes the study of the MLA-DA model in the cross-domain fault diagnosis of induction motors.

參考文獻
[1] X. Liang, M. Z. Ali and H. Zhang, "Induction Motors Fault Diagnosis Using Finite Element Method: A Review," IEEE Transactions on Industry Applications, vol. 56, no. 2, pp. 1205-1217, 2020.
[2] Y. Lei, B. Yang, X. Jiang, F. Jia, N. Li, A. K. Nandi,” Applications of Machine Learning to Machine Fault Diagnosis: A Review and Roadmap,” Mechanical Systems and Signal Processing, vol. 138, 2020.
[3] B. Liu, H. Pan and X. Li, "An Expert System for Fault Diagnosis in Diesel Engine Based on Wavelet Packet Analysis and Hybrid PSO-DV Based Neural Network," 2010 International Conference on Intelligent Computing and Cognitive Informatics, Kuala Lumpur, Malaysia, pp. 29-32, 2010.
[4] Y. Wang, "Research on the Fault Diagnosis of Mechanical Equipment Vibration System Based on Expert System," 2018 International Conference on Sensing, Diagnostics, Prognostics, and Control (SDPC), Xi'an, pp. 636-641, China, 2018.
[5] J. F. Martins, V. Ferno Pires, A. J. Pires, "Unsupervised Neural-Network-Based Algorithm for an On-Line Diagnosis of Three-Phase Induction Motor Stator Fault," IEEE Transactions on Industrial Electronics, vol. 54, no. 1, pp. 259-264, 2007.
[6] T. dos Santos, F. J. T. E. Ferreira, J. M. Pires, C. Damásio, "Stator Winding Short-circuit Fault Diagnosis in Induction Motors Using Random Forest," 2017 IEEE International Electric Machines and Drives Conference (IEMDC), Miami, FL, USA, pp. 1-8, 2017.
[7] Adam Glowacz, “Acoustic Based Fault Diagnosis of Three-phase Induction Motor,” Applied Acoustics, vol. 137, pp. 82-89, 2018.
[8] M. Z. Ali, M. N. S. K. Shabbir, X. Liang, Y. Zhang, T. Hu, "Machine Learning-Based Fault Diagnosis for Single- and Multi-Faults in Induction Motors Using Measured Stator Currents and Vibration Signals," IEEE Transactions on Industry Applications, vol. 55, no. 3, pp. 2378-2391, 2019.
[9] K. Yatsugi, S. E. Pandarakone, Y. Mizuno and H. Nakamura, "Common Diagnosis Approach to Three-Class Induction Motor Faults Using Stator Current Feature and Support Vector Machine," IEEE Access, vol. 11, pp. 24945-24952, 2023.
[10] Czeslaw T Kowalski, Teresa Orlowska-Kowalska, “Neural Networks Application for Induction Motor Faults Diagnosis,” Mathematics and Computers in Simulation, Vol. 63, pp. 435-448, 2003.
[11] X. Lin, B. Li, X. Yang and J. Wang, "Fault Diagnosis of Aero-engine Bearing Using a Stacked Auto-Encoder Network," 2018 IEEE 4th Information Technology and Mechatronics Engineering Conference (ITOEC), Chongqing, China, pp. 545-548, 2018.
[12] Y. Li, L. Zou, L. Jiang and X. Zhou, "Fault Diagnosis of Rotating Machinery Based on Combination of Deep Belief Network and One-dimensional Convolutional Neural Network," IEEE Access, vol. 7, pp. 165710-165723, 2019.
[13] J. Guo and P. Zheng, "A Method of Rolling Bearing Fault Diagnose Based on Double Sparse Dictionary and Deep Belief Network," IEEE Access, vol. 8, pp. 116239-116253, 2020.
[14] L. Wen, X. Li, L. Gao and Y. Zhang, "A New Convolutional Neural Network-Based Data-Driven Fault Diagnosis Method," IEEE Transactions on Industrial Electronics, vol. 65, no. 7, pp. 5990-5998, 2018.
[15] L. Yuan, D. Lian, X. Kang, Y. Chen and K. Zhai, "Rolling Bearing Fault Diagnosis Based on Convolutional Neural Network and Support Vector Machine," IEEE Access, vol. 8, pp. 137395-137406, 2020.
[16] S. Shao, R. Yan, Y. Lu, P. Wang and R. X. Gao, "DCNN-Based Multi-Signal Induction Motor Fault Diagnosis," IEEE Transactions on Instrumentation and Measurement, vol. 69, no. 6, pp. 2658-2669, 2020.
[17] J. Yang, J. Liu, J. Xie, C. Wang and T. Ding, "Conditional GAN and 2-D CNN for Bearing Fault Diagnosis with Small Samples," IEEE Transactions on Instrumentation and Measurement, vol. 70, pp. 1-12, Art no. 3525712, 2021.
[18] L. Hou, R. Jiang, Y. Tan and J. Zhang, "Input Feature Mappings-Based Deep Residual Networks for Fault Diagnosis of Rolling Element Bearing with Complicated Dataset," IEEE Access, vol. 8, pp. 180967-180976, 2020.
[19] W. Li, R. Huang, J. Li, Y. Liao, Z. Chen, G. He, R. Yan, Konstantinos Gryllias, “A Perspective Survey on Deep Transfer Learning for Fault Diagnosis in Industrial Scenarios: Theories, Applications and Challenges,” Mechanical Systems and Signal Processing, vol. 167, 2022.
[20] Z. Zhang, H. Chen, S. Li, Z. An, “Unsupervised Domain Adaptation via Enhanced Transfer Joint Matching for Bearing Fault Diagnosis,” Measurement, vol. 165, 2020.
[21] B. Yang, Y. Lei, S. Xu and C. G. Lee, "An Optimal Transport-Embedded Similarity Measure for Diagnostic Knowledge Transferability Analytics Across Machines," IEEE Transactions on Industrial Electronics, vol. 69, no. 7, pp. 7372-7382, 2022.
[22] S. Shao, S. McAleer, R. Yan and P. Baldi, "Highly Accurate Machine Fault Diagnosis Using Deep Transfer Learning," IEEE Transactions on Industrial Informatics, vol. 15, no. 4, pp. 2446-2455, 2019.
[23] Z. Chen, J. Cen and J. Xiong, "Rolling Bearing Fault Diagnosis Using Time-Frequency Analysis and Deep Transfer Convolutional Neural Network," IEEE Access, vol. 8, pp. 150248-150261, 2020.
[24] D. Xiao, Y. Huang, L. Zhao, C. Qin, H. Shi and C. Liu, "Domain Adaptive Motor Fault Diagnosis Using Deep Transfer Learning," IEEE Access, vol. 7, pp. 80937-80949, 2019.
[25] W. Zhang, P. Zhang, X. He and D. Zhang, "Convolutional Neural Network Based Two-Layer Transfer Learning for Bearing Fault Diagnosis," IEEE Access, vol. 10, pp. 109779-109794, 2022.
[26] J. Zhu, N. Chen and C. Shen, "A New Deep Transfer Learning Method for Bearing Fault Diagnosis Under Different Working Conditions," IEEE Sensors Journal, vol. 20, no. 15, pp. 8394-8402, 2020.
[27] Z. Wang, X. He, B. Yang and N. Li, "Subdomain Adaptation Transfer Learning Network for Fault Diagnosis of Roller Bearings," IEEE Transactions on Industrial Electronics, vol. 69, no. 8, pp. 8430-8439, 2022.
[28] Y. Wang, X. Sun, J. Li and Y. Yang, "Intelligent Fault Diagnosis With Deep Adversarial Domain Adaptation," IEEE Transactions on Instrumentation and Measurement, vol. 70, pp. 1-9, 2021, Art no. 2503509.
[29] Q. Li, S. Liu, B. Yang, Y. Xu, L. Chen and C. Shen, "Adversarial Multi-Domain Adaptation for Machine Fault Diagnosis with Variable Working Conditions," 2020 IEEE 18th International Conference on Industrial Informatics (INDIN), Warwick, United Kingdom, pp. 737-741, 2020.
[30] S. Pang and X. Yang, "A Cross-Domain Stacked Denoising Autoencoders for Rotating Machinery Fault Diagnosis Under Different Working Conditions," IEEE Access, vol. 7, pp. 77277-77292, 2019.
[31] Z. -H. Liu, L. -B. Jiang, H. -L. Wei, L. Chen and X. -H. Li, "Optimal Transport-Based Deep Domain Adaptation Approach for Fault Diagnosis of Rotating Machine," IEEE Transactions on Instrumentation and Measurement, vol. 70, pp. 1-12, Art no. 3508912, 2021.
[32] H. Zhong, H. Tuo, C. Wang, X. Ren, J. Hu and L. Qiao, "Source-Constraint Adversarial Domain Adaptation," 2019 IEEE International Conference on Image Processing (ICIP), pp. 2486-2490, Taipei, Taiwan, 2019.
[33] X.D. Wang, F. Liu, “Triplet Loss Guided Adversarial Domain Adaptation for Bearing Fault Diagnosis”, Sensors, vol. 20, no. 1, 2020.
[34] J. Ryu, M.H. Yang, J. Lim, “Unsupervised Feature Learning for Self-tuning Neural Networks,” Neural Networks, Vol. 133, pp. 103-111, 2021.
[35] B. Geng, D. Tao and C. Xu, "DAML: Domain Adaptation Metric Learning," IEEE Transactions on Image Processing, vol. 20, no. 10, pp. 2980-2989, 2011.
[36] X. Liang, M. Z. Ali and H. Zhang, "Induction Motors Fault Diagnosis Using Finite Element Method: A Review," IEEE Transactions on Industry Applications, vol. 56, no. 2, pp. 1205-1217, 2020.
[37] A. H. Bonnett, C. Yung, "Increased Efficiency Versus Increased Reliability," IEEE Industry Applications Magazine, vol. 14, no. 1, pp. 29-36, 2008.
[38] J. L. Lin, Y. C. Liu, C. W. Li, L. F. Tsai, H. Y. Chung, "Motor Shaft Misalignment Detection Using Multiscale Entropy with Wavelet Denoising," Expert Systems with Applications, vol. 37, no. 10, pp. 7200-7204, 2010.
[39] National Instrument, “NI PXI-8115 Intel Core i5-2510EDual-CorePXI Controller” June 2013, http://sine.ni.com/nips/cds/view/p/lang/en/nid/204583
[40] National Instrument, “NI PXIe-4300DataAcquisitionwith Integrated Signal Conditioning for High-Voltage Measurements” June 2013, http://sine.ni.com/ds/app/doc/p/id/ds-207/lang/zht.
[41] PCB Piezotronics, “PCB 352C03 Accelerometer, ICP”, https://www.pcb.com/products?m=352C03.
[42] L. Wen, X. Li, L. Gao, Y. Zhang, "A New Convolutional Neural Network Based Data-Driven Fault Diagnosis Method," IEEE Transection Industrial Electronic, vol. 65, no. 7, pp. 5990-5998, 2018.
[43] E. Tzeng, J. Hoffman, N. Zhang, K. Saenko, T. Darrell, "Deep Domain Confusion: Maximizing for Domain Invariance," arXiv:1412.3474, 2014.
[44] A. Gretton, K.M. Borgwardt, M.J. Rasch, B. Schoelkopf, Bernhard, A. Smola, "A Kernel Two-Sample Test, " Journal of Machine Learning Research, vol. 13, pp. 723-773, 2012.
[45] Y. Ganin et al., "Domain-Adversarial Training of Neural Networks, " Journal of Machine Learning Research, vol. 17, p. 1-35, 2016.
[46] Y. Zhu et al., "Deep Subdomain Adaptation Network for Image Classification," IEEE Transactions on Neural Networks and Learning Systems, vol. 32, no. 4, pp. 1713-1722, 2021.
[47] He, X. Zhang, S. Ren, J. Sun, "Deep Residual Learning for Image Recognition", Proc. IEEE Conference Compute Visual Pattern Recognition (CVPR), pp. 770-778, 2016.
[48] Y.M. Yin, Z. Yang, H.F. Hu, X.F. Wu, ”Metric Learning-Assisted Domain Adaptation,” Neurocomputing, vol. 454, pp. 268-279, 2021.
[49] F. Schroff, D. Kalenichenko and J. Philbin, "FaceNet: A Unified Embedding for Face Recognition and Clustering," 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 815-823, Boston, MA, USA, 2015.
[50] S. Zhang, Q. Zhang, X. Wei, Y. Zhang and Y. Xia, "Person Re-Identification with Triplet Focal Loss," IEEE Access, vol. 6, pp. 78092-78099, 2018.
[51] Y. Li, G. Yin, C. Liu, Xi. Yang, Z. Wang, “Triplet Online Instance Matching Loss for Person Re-identification,” Neurocomputing, vol. 433, pp. 10-18, 2021.
[52] W. Xie, H. Wu, Y. Tian, M. Bai and L. Shen, "Triplet Loss with Multistage Outlier Suppression and Class-Pair Margins for Facial Expression Recognition," IEEE Transactions on Circuits and Systems for Video Technology, vol. 32, no. 2, pp. 690-703, 2022.
[53] B.H. Chen, W.H. Deng, “Deep Embedding Learning with Adaptive Large Margin N-pair Loss for Image Retrieval and Clustering,” Pattern Recognition, Vol. 93, pp. 353-364, 2019.
[54] Baeza-Yates, Ricardo, and Berthier Ribeiro-Neto. Modern Information Retrieval. 2nd ed., Addison-Wesley, 2011.
[55] Sasaki, Yutaka. "The Truth of the F-Measure." Journal of Machine Learning Research, vol. 8, pp. 1-24, 2007.