本研究旨在比較深度域混淆網路(Deep Domain Confuse, DDC) 、域對抗神經網路(Domain Adversarial Neural Network, DANN) 與子域適應網路(Deep Subdomain Adaptation Network, DSAN) 等三種域適應網路(Domain-Adaptation) 對三相感應馬達跨域故障(Cross-Domain)診斷之成效，包含不同負載之間與不同容量之間的跨域診斷，並透過案例設計與結果分析，設計一套具綜合跨負載與跨容量之感應馬達跨域故障診斷系統。
首先，選定2馬力馬達與5馬力馬達作為本研究跨容量目標，分別設計定子故障、轉子故障、軸承故障、不對心故障與健康狀態等五種馬達實驗模型，並量測其於不同負載(0%載、25%載、50%載、75%載與100%載) 下的振動訊號，再將原始時域訊號轉為圖像檔並整理為圖像集，以供後續域適應網路模型訓練使用。其次，以2馬力之五種負載的振動訊號圖像集分別作為源域(Source-Domain) 與目標域(Target-Domain)，匯入三種域適應網路並分別訓練，除了比較不同域適應網路對馬達跨負載故障診斷之成效，亦比較以何種負載作為源域進行訓練可得到最佳的準確率。再者，以相同負載之2馬力與5馬力的振動訊號圖像集，分別作為源域與目標域，匯入三種域適應網路並加以訓練，並根據其準確率，驗證跨容量故障診斷之可行性。最後，根據跨負載故障診斷之結果顯示，以50%載作為源域訓練可達到99%之診斷準確率，因此以2馬力之50%載振動訊號圖像集作為源域，5馬力各負載之振動訊號圖像集分別作為目標域，匯入域適應網路並訓練，藉此達成綜合跨負載與跨容量之馬達故障診斷，結果顯示其診斷準確率仍可達到82%，足以驗證綜合跨負載與跨容量故障診斷之可行性。

This study aims to compare three kinds of domain adaptation networks' effectiveness in a cross-domain diagnosis of three-phase induction motors, including Deep Domain Confuse(DDC), Domain Adversarial Neural Network (DANN), and Deep Subdomain Adaptation Network (DSAN). Through case design and result in analysis, to achieve fault diagnosis across loads and capacities of motors, and design an induction motor cross-domain diagnosis system comprehensive with cross-load and cross-capacity fault diagnosis.
First, 2-horsepower motors and 5-horsepower motors are selected as the cross-capacity diagnosis targets of this research, and five motor experimental models, including stator fault, rotor fault, bearing fault, misalignment fault, and general health state, are designed respectively, and they are measured under different loads ( 0% load, 25% load, 50% load, 75% load, and 100% load) to get their vibration signal, and then convert the original time-domain signal into an image file and organize it into an image set for use of subsequent domain adaptation network Model training. Secondly, the vibration signal image sets of five loads of 2 horsepower motor are respectively used as the source domain and the target domain, import the three domain adaptation networks and train them, to compare different domain adaptation networks' effectiveness in the cross-load fault diagnosis and also compared which loads are used as the source domain for training can obtain the best accuracy. Furthermore, the vibration signal image sets of 2 horsepower and 5 horsepower motors with the same load are used as the source domain and the target domain, respectively, import the three domain adaptation networks and train them. Its accuracy verifies the feasibility of cross-capacity fault diagnosis. Finally, according to the results of cross-load fault diagnosis, the model which trains with 50% load as the source domain can obtain better accuracy of fault diagnosis. Therefore, the vibration signal image set of 50% load of 2 horsepower is used as the source domain, and the vibration signal of each load of 5 horsepower is used as the target domain. which are imported into the domain adaptation network and trained to achieve comprehensive cross-load and cross-capacity motor fault diagnosis.

[1] C. N. Jibhakate, M. A. Chaudhari, and M. M. Renge, ‘‘Reactive Power
Compensation Using Induction Motor Driven by Nine Switch AC-DC-AC
Converter,’’ IEEE Access, vol. 6, pp. 1312-1320, 2018.
[2] M. A. Hannan et al., ‘‘A Quantum Lightning Search Algorithm-based
Fuzzy Speed Controller for Induction Motor Drive,’’ IEEE Access, vol. 6,
pp. 1214–1223, 2018.
[3] F. Giri, AC Electric Motors Control: Advanced Design Techniques and
Applications. Hoboken, NJ, USA: Wiley, 2013.
[4] 曾思憲，「感應馬達主動性狀態估測系統之研發」，碩士論文，臺灣
科技大學，2018 年。
[5] S. Nandi, H. A. Toliyat and X. Li, “Condition Monitoring and Fault
Diagnosis of Electrical Motors—A Review,” IEEE Trans. Energy
Convers., vol. 20, no. 4, pp. 719-729, Dec. 2005.
[6] 林承憲，「應用域對抗神經網路於感應馬達跨域故障診斷」，碩士論
文，臺灣科技大學，2020 年。
[7] D. Neupane and J. Seok, “Bearing Fault Detection and Diagnosis Using
Case Western Reserve University Dataset With Deep Learning
Approaches: A Review,” IEEE Access, vol. 8, pp. 93155-93178, 2020.
[8] D. T. Hoang, H.J. Kang, “A Survey on Deep Learning Based Bearing
Fault Diagnosis,” Neurocomputing, Vol 335, pp 327-335, 2019.
[9] Case Western Reserve University Bearing Data Center Website. 2000.
[Online]. Available: http://csegroups.case.edu/bearingdatacenter/home.
80
[10] H. Qiu, J. Lee, J. Lin, and G. Yu, “Wavelet Filter-Based Weak Signature
Detection Method and Its Application on Rolling Element Bearing
Prognostics,” Journal of Sound and Vibration, vol. 289, no. 4, pp. 1066-
1090, 2006.
[11] C. Zhang, L. Xu, X. Li and H. Wang, “A Method of Fault diagnosis for
rotary equipment based on deep learning,” Prognostics and System Health
Management Conf. (PHM-Chongqing), 2018, pp. 958-962.
[12] L. Wen, X. Li, L. Gao and Y. Zhang, “A New Convolutional Neural
Network Based Data-Driven Fault Diagnosis Method,” IEEE Trans. Ind.
Electron, vol. 65, no. 7, pp. 5990-5998, Jul. 2018.
[13] 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.
[14] Q. Liu and C. Huang, “A Fault Diagnosis Method Based on Transfer
Convolutional Neural Networks,” IEEE Access, vol. 7, pp. 171423-
171430, 2019.
[15] W. Xu, Y. Wan, T.Y. Zuo and X.M. Sha, “Transfer Learning Based Data
Feature Transfer for Fault Diagnosis,” IEEE Access, vol. 8, pp. 76120-
76129, 2020.
[16] Y. Zhou, Y. Dong, H. Zhou and G. Tang, “Deep Dynamic Adaptive
Transfer Network for Rolling Bearing Fault Diagnosis With Considering
Cross-Machine Instance,” IEEE Trans. Instrumentation and Measurement,
vol. 70, pp. 1-11, 2021.
[17] Y. Lei, B. Yang, X. Jiang, F. Jia, N. Li, and A. K. Nandi, “Applications of
Machine Learning to Machine Fault Diagnosis: A Review and Roadmap,”
Mech. Syst. Signal Process, vol. 138, 2020.
81
[18] L. Y. Pratt et al, “Discriminability-Based Transfer Between Neural
Networks,” Proc. Int. Conf. Neural Inf. Process. Syst, 1992, pp 204-211.
[19] S. J. Pan and Q. Yang, "A Survey on Transfer Learning," IEEE Trans.
Knowledge and Data Engineering, vol. 22, no. 10, pp. 1345-1359, 2010.
[20] L. Mihalkova, T. Huynh, and R.J. Mooney, “Mapping and Revising
Markov Logic Networks for Transfer Learning,” Proc. 22nd AAAI. Conf.
Artificial Intelligence, 2007, pp. 608-614.
[21] R. Raina, A. Battle, H. Lee, B. Packer, and A.Y. Ng, “Self-Taught Learning:
Transfer Learning from Unlabeled Data,” Proc. 24th Int’l Conf. Machine
Learning, 2007, pp. 759-766.
[22] E. Eaton, M. desJardins, and T. Lane, “Modeling Transfer Relationships
between Learning Tasks for Improved Inductive Transfer,” Proc.
European Conf. Machine Learning and Knowledge Discovery in
Databases (ECML/PKDD ’08), 2008, pp. 317-332.
[23] E. Tzeng, J. Hoffman, N. Zhang, K. Saenko and T. Darrell, "Deep domain
confusion: Maximizing for domain invariance," Journal of E-point, pp
1412-1417, 2014.
[24] Y. Ganin et al., “Domain-Adversarial Training of Neural Networks,” J.
Mach. Learn. Res., vol. 17, no. 1, pp. 2030–2096, 2016.
[25] Y. Zhu et al., “Deep Subdomain Adaptation Network for Image
Classification,” IEEE Trans. Neural Networks and Learning Systems, vol.
32, no. 4, pp. 1713-1722, 2022.
[26] Hu, R., Zhang, M., Xiang, Z. “Guided Deep Subdomain Adaptation
Network for Fault Diagnosis of Different Types of Rolling Bearings,”
Journal of Intell Manuf., 2022.
82
[27] Agam G., Gurmeet Singh, and V.N.A. Naikan, “Effective Combination of
Motor Fault Diagnosis Techniques,” 2018 Int. Conf. on Power,
Instrumentation, Control and Computing (PICC), 2018, pp 1-5.
[28] 林芝以，「運用電氣訊號於感應馬達狀態監測與故障診斷之研究」，
碩士論文，台灣科技大學電機工程學系，2016 年。
[29] PCB Piezotronics, “PCB 352C03 Accelerometer, ICP”,
https://www.pcb.com/products?m=352C03.
[30] 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.
[31] 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.
[32] D. Kifer, S. Ben-David, and J. Gehrke. “Detecting Change in Data
Streams,” Very Large Data Bases, pp 180-191, 2004.
[33] S. Ben-David, J. Blitzer, K. Crammer, and F. Pereira, “Analysis of
Representations for Domain Adaptation,” Proc. NIPS, conf., 2007, pp.
137-144.
[34] M. Yishay, M. Mehryar, and R. Afshin. “Domain Adaptation: Learning
Bounds and Algorithms,” COLT, Conf. 2009.
[35] B.D. Shai, B. John, C. Koby, K. Alex, P. Fernando, and W. V. Jennifer.
“A Theory of Learning From Different Domains,” Machine Learning,
79(1-2), pp 151-175, 2010.
83
[36] P. Germain, A. Habrard, C Laviolette, and E. Morvant. “A PACBayesian
Approach for Domain Adaptation with Specialization to Linear
Classifiers,” ICML, pp 738-746, 2013.
[37] K. He, X. Zhang, S. Ren, and J. Sun, “Deep Residual Learning for Image
Recognition,” Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR),
2016, pp. 770-778.