在預測和健康管理（prognostic and health management, PHM）中，剩餘使用壽命（remaining useful life, RUL）預測是關鍵任務之一。然而，在某些特定機器上獲得完整的運行到故障記錄不易。為了解決這個問題，讓新機器可以參考相同或類似類型的其他機器來開發預測模型，遷移學習(transfer learning)的方法被應用在實務之中，其中「領域適應(domain adaptation, DA)」的方法在近年來是熱門的研究之一，領域適應是一種用於將從源域獲得的知識遷移到目標域的方法。本研究基於領域適應的概念提出了特徵增強的多源子域自適應 (feature-enhanced multi-source subdomain adaptation, FEMSA) 來預測 RUL。 FEMSA透過重新定義多個源域來學習域不變特徵並表徵相似性；也就是說，我們透過重新制定多個操作條件來處理跨域泛化，另一方面，本研究同時考慮當經過一段時間之後，資料可能與原本的分配不一致，因此加入飄移偵測(concept drift)的機制，以提供穩定的模型預測結果。最後更將RUL結合保養排程，利用數學規劃模型求解，合理安排機台保養，使產能損失降低。此外，應用兩個公開資料集來驗證所提出的 FEMSA，結果顯示 FEMSA 可以提供更穩健的 RUL 預測，而飄移偵測的機制在模擬的情境中，隨著時間改變，預測表現也能維持平穩，綜上所述，本研究提出的架構能夠穩健的預測剩餘壽命並改進 PHM 系統。

In prognostic and health management (PHM), the remaining useful life (RUL) prediction is one of the key tasks. However, a complete run-to-failure record (with label) is not always available on some specific machines, for example, the new machine. To address the issue, the new machine may refer to other machines of the same or similar type (even old machines with labels) for developing the prediction model. Transfer learning methods have been applied in practice, particularly domain adaptation, which methodology used to transfer the knowledge gained from the source domain to the target domain, is one of the most popular research areas in recent years. This study proposes feature-enhanced multi-source subdomain adaptation (FEMSA) to predict the RUL based on the domain adaptation. FEMSA learns the domain-invariant features and characterizes the similarity by redefining the multiple source domains; that is, we handle the cross-domain generalization by reformulating the multiple operating conditions. On the other hand, considering that after a certain period, the data may not be consistent with the original assignment, and drift detection mechanism is added. Finally, RUL is combined with maintenance schedules, using mathematical programming to solve problems, and rationally arrange machine maintenance to reduce production yield losses. In the experimental study, the result shows that the FEMSA can provide more robust RUL prediction over time. Furthermore, the drift detection mechanism can maintain a stable prediction performance with the change of time in the simulated scenarios. In conclusion, the proposed framework can provide a robust prediction of the RUL and improve the PHM system.

[1] Lu, H. W., & Lee, C. Y., “Kernel-based dynamic ensemble technique for remaining useful life prediction,” IEEE Robotics and Automation Letters, 7(2), 1142-1149, 2022.
[2] Lu, H. W., & Lee, C. Y., “Feature-Enhanced Multisource Subdomain Adaptation on Robust Remaining Useful Life Prediction,” IEEE Robotics and Automation Letters, 9(7), 6130-6137, 2024
[3] Baheti, R., & Gill, H. Cyber-physical systems. The impact of control technology, 12(1), 161-166, 2011.
[4] Lee, J., Bagheri, B., & Kao, H. A. A cyber-physical systems architecture for industry 4.0-based manufacturing systems. Manufacturing letters, 3, 18-23, 2015.
[5] Lee, J., & Kundu, P. Integrated cyber-physical systems and industrial metaverse for remote manufacturing. Manufacturing Letters, 34, 12-15, 2022.
[6] Dafflon, B., Moalla, N., & Ouzrout, Y. The challenges, approaches, and used techniques of CPS for manufacturing in Industry 4.0: a literature review. The International Journal of Advanced Manufacturing Technology, 113, 2395-2412, 2021.
[7] Paelke V, R¨ocker C, Koch N, Flatt H, B¨uttner S. User interfaces for cyber-physical systems. at-Automatisierungstechnik 63(10), 833-843, 2015.
[8] Pirvu B-C, Zamfirescu C-B, Gorecky D. Engineering insights from an anthropocentric cyber-physical system: a case study for an assembly station. Mechatronics, 34, 147-159, 2016.
[9] Zhao, Z., Liang, B., Wang, X., & Lu, W. “Remaining useful life prediction of aircraft engine based on degradation pattern learning,” Reliability Engineering & System Safety, 164, 74-83, 2017.
[10] Pan, S. J., & Yang, Q., “A survey on transfer learning,” IEEE Transactions on knowledge and data engineering, 22(10), 1345-1359, 2010.

[11] Pan, Z., Meng, Z., Chen, Z., Gao, W., & Shi, Y., “A two-stage method based on extreme learning machine for predicting the remaining useful life of rolling-element bearings,” Mechanical Systems and Signal Processing, 144, 106899, 2020.
[12] Cubillo, A., Perinpanayagam, S., & Esperon-Miguez, M., “A review of physics-based models in prognostics: Application to gears and bearings of rotating machinery,” Advances in Mechanical Engineering, 8(8), 1687814016664660, 2016.
[13] Kundu, P., Darpe, A. K., & Kulkarni, M. S., “An ensemble decision tree methodology for remaining useful life prediction of spur gears under natural pitting progression,” Structural Health Monitoring, 19(3), 854-872, 2020.
[14] Xia, M., Li, T., Shu, T., Wan, J., De Silva, C. W., & Wang, Z., “A two-stage approach for the remaining useful life prediction of bearings using deep neural networks,” IEEE Transactions on Industrial Informatics, 15(6), 3703-3711, 2019.
[15] Liu, R., Yang, B., & Hauptmann, A. G., “Simultaneous bearing fault recognition and remaining useful life prediction using joint-loss convolutional neural network,” IEEE Transactions on industrial informatics, 16(1), 87-96, 2019.
[16] Y. Wu, M. Yuan, S. Dong, L. Lin, and Y. Liu, “Remaining useful life estimation of engineered systems using vanilla LSTM neural networks,” Neurocomputing, 275, 167-179, 2018.
[17] Kim, Y. H., Shin, U., Park, J., & Kweon, I. S. MS-UDA: Multi-spectral unsupervised domain adaptation for thermal image semantic segmentation. IEEE Robotics and Automation Letters, 6(4), 6497-6504, 2021.
[18] Zhang, J., Tai, L., Yun, P., Xiong, Y., Liu, M., Boedecker, J., & Burgard, W. Vr-goggles for robots: Real-to-sim domain adaptation for visual control. IEEE Robotics and Automation Letters, 4(2), 1148-1155, 2019.
[19] Zheng, L., Ma, W., Cai, Y., Lu, T., & Wang, S. GPDAN: Grasp Pose Domain Adaptation Network for Sim-to-Real 6-DoF Object Grasping. IEEE Robotics and Automation Letters, 2023.
[20] Ganin, Y., Ustinova, E., Ajakan, H., Germain, P., Larochelle, H., Laviolette, F., Marchand, M., & Lempitsky, V., “Domain-adversarial training of neural networks,” Journal of Machine Learning Research, 17(1), 2096-2030, 2016.
[21] da Costa, P. R. D. O., Akçay, A., Zhang, Y., & Kaymak, U., “Remaining useful lifetime prediction via deep domain adaptation,” Reliability Engineering & System Safety, 195, 106682, 2020.
[22] Cheng, H., Kong, X., Chen, G., Wang, Q., & Wang, R., “Transferable convolutional neural network based remaining useful life prediction of bearing under multiple failure behaviors,” Measurement, 168, 108286, 2021.
[23] Ragab, M., Chen, Z., Wu, M., Foo, C. S., Kwoh, C. K., Yan, R., & Li, X., “Contrastive adversarial domain adaptation for machine remaining useful life prediction,” IEEE Transactions on Industrial Informatics, 17(8), 5239-5249, 2020.
[24] R. Xu, Z. Chen, W. Zuo, J. Yan, and L. Lin, “Deep cocktail network: Multisource unsupervised domain adaptation with category shift,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., 3964–3973, 2018.
[25] Y. Zhu, F. Zhuang, and D. Wang, “Aligning domain-specific distribution and classifier for cross-domain classification from multiple sources,” Assoc. Adv. Artif. Intell., 33(1), 5989–5996, Jul. 2019.
[26] N. Venkat, J. N. Kundu, D. Singh, A. Revanur, and V. B. R., “Your classifier can secretly suffice multi-source domain adaptation,” in Proc. Adv. Neural Inf. Process. Syst., 4647–4659, 2020.
[27] Gretton, A., Borgwardt, K. M., Rasch, M. J., Schölkopf, B., & Smola, A., “A kernel two-sample test, “Journal of Machine Learning Research, 13(1), 723-773, 2012.
[28] Rüschendorf, L., “The Wasserstein distance and approximation theorems,“ Probability Theory and Related Fields, 70(1), 117-129, 1985.
[29] Sun, B., & Saenko, K. Deep coral: Correlation alignment for deep domain adaptation. In Computer Vision–ECCV 2016 Workshops: Amsterdam, The Netherlands, October 8-10 and 15-16, 2016, Proceedings, Part III 14 (443-450). Springer International Publishing, 2016.
[30] Rezaeianjouybari, B., & Shang, Y., “A novel deep multi-source domain adaptation framework for bearing fault diagnosis based on feature-level and task-specific distribution alignment,” Measurement, 178, 109359, 2021

[31] Bonneel, N., Rabin, J., Peyré, G., & Pfister, H., “Sliced and radon wasserstein barycenters of measures,” Journal of Mathematical Imaging and Vision, 51(1), 22-45, 2015.
[32] Xia, Y., Shen, C., Wang, D., Shen, Y., Huang, W., & Zhu, Z., “Moment matching-based intraclass multisource domain adaptation network for bearing fault diagnosis,” Mechanical Systems and Signal Processing, 168, 108697, 2022.
[33] Zhu, Y., Zhuang, F., Wang, J., Ke, G., Chen, J., Bian, J., Xiong, H., & He, Q., “Deep subdomain adaptation network for image classification,” IEEE Transactions on Neural Networks and Learning Systems, 32(4), 1713-1722, 2020.
[34] Liu, Y., Wang, Y., Chow, T. W., & Li, B. Deep adversarial subdomain adaptation network for intelligent fault diagnosis. IEEE Transactions on Industrial Informatics, 18(9), 6038-6046, 2022.
[35] Wei, P., Ke, Y., Qu, X., & Leong, T. Y., “Subdomain adaptation with manifolds discrepancy alignment,” IEEE Transactions on Cybernetics, 52(11), 11698-11708, 2021.
[36] Tian, J., Han, D., Li, M., & Shi, P. A multi-source information transfer learning method with subdomain adaptation for cross-domain fault diagnosis. Knowledge-Based Systems, 243, 108466, 2022.
[37] Tan, C., Sun, F., Kong, T., Zhang, W., Yang, C., & Liu, C. A survey on deep transfer learning. In Artificial Neural Networks and Machine Learning–ICANN 2018: 27th International Conference on Artificial Neural Networks, Rhodes, Greece, October 4-7, 2018, Proceedings, Part III 27 (pp. 270-279). Springer International Publishing, 2018.
[38] Xie, J., Zhang, L., Duan, L., & Wang, J. “On cross-domain feature fusion in gearbox fault diagnosis under various operating conditions based on transfer component analysis.” In 2016 IEEE international conference on prognostics and health management (ICPHM) (pp. 1-6). IEEE, 2016
[39] Lu, W., Liang, B., Cheng, Y., Meng, D., Yang, J., & Zhang, T. “Deep model based domain adaptation for fault diagnosis”. IEEE Transactions on Industrial Electronics, 64(3), 2296-2305, 2016

[40] Li X, Zhang W, Ding Q. Cross-domain fault diagnosis of rolling element bearings using deep generative neural networks. IEEE Transactions on Industrial Electronics, 66(7), 5525-5534, 2018.
[41] Wen, L., Gao, L., & Li, X., “A new deep transfer learning based on sparse auto-encoder for fault diagnosis,” IEEE Transactions on Systems, Man, and Cybernetics: Systems, 49(1), 136-144, 2017.
[42] Ding, Y., Jia, M., Miao, Q., & Huang, P., “Remaining useful life estimation using deep metric transfer learning for kernel regression,” Reliability Engineering & System Safety, 212, 107583, 2021.
[43] Ding, Y., Jia, M., & Cao, Y., “Remaining useful life estimation under multiple operating conditions via deep subdomain adaptation,” IEEE Transactions on Instrumentation and Measurement, 70, 1-11, 2021.
[44] Zhang, J., Li, X., Tian, J., Jiang, Y., Luo, H., & Yin, S. A variational local weighted deep sub-domain adaptation network for remaining useful life prediction facing cross-domain condition. Reliability Engineering & System Safety, 231, 108986, 2023.
[45] Hu, T., Guo, Y., Gu, L., Zhou, Y., Zhang, Z., & Zhou, Z., “Remaining useful life estimation of bearings under different working conditions via Wasserstein distance-based weighted domain adaptation,” Reliability Engineering & System Safety, 224, 108526, 2022.
[46] Agrahari, S., & Singh, A. K., Concept drift detection in data stream mining: A literature review. Journal of King Saud University-Computer and Information Sciences, 34(10), 9523-9540, 2022.
[47] Li, P., Wu, X., Hu, X., & Wang, H., Learning concept-drifting data streams with random ensemble decision trees. Neurocomputing, 166, 68-83, 2015.
[48] Gözüaçık, Ö., & Can, F., Concept learning using one-class classifiers for implicit drift detection in evolving data streams. Artificial Intelligence Review, 54(5), 3725-3747, 2021.
[49] Liu, A., Zhang, G., & Lu, J. Fuzzy time windowing for gradual concept drift adaptation. In 2017 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE) (pp. 1-6). IEEE, 2017.
[50] Lee, C.-Y., & Chien, C.-F., “Pitfalls and protocols of data science in manufacturing practice,” J. Intell. Manuf., 33, 1189–1207, 2022.
[51] Jiang, W., Hong, Y., Zhou, B., He, X., & Cheng, C., “A GAN-based anomaly detection approach for imbalanced industrial time series,” IEEE Access, 7, 143608-143619, 2019.
[52] Lee, C. Y., Huang, T. S., Liu, M. K., & Lan, C. Y., “Data science for vibration heteroscedasticity and predictive maintenance of rotary bearings,” Energies, 12(5), 801, 2019.
[53] Wen, L., Gao, L., & Li, X., “A new deep transfer learning based on sparse auto-encoder for fault diagnosis,” IEEE Transactions on Systems, Man, and Cybernetics: Systems, 49(1), 136-144, 2017
[54] Ren, L., Sun, Y., Cui, J., & Zhang, L., “Bearing remaining useful life prediction based on deep autoencoder and deep neural networks,” Journal of Manufacturing Systems, 48, 71-77, 2018.
[55] Cheng, Z., Sun, H., Takeuchi, M., & Katto, J., “Deep convolutional autoencoder-based lossy image compression,” 2018 Picture Coding Symposium (PCS), 253-257, 2018.
[56] Masud, M. A., Huang, J. Z., Wei, C., Wang, J., Khan, I., & Zhong, M., “I-nice: A new approach for identifying the number of clusters and initial cluster centres,” Information Sciences, 466, 129-151, 2018.
[57] Dempster, A. P., Laird, N. M., & Rubin, D. B., “Maximum likelihood from incomplete data via the EM algorithm,” Journal of the Royal Statistical Society: Series B, 39(1), 1-22, 1977.
[58] Lee, C.-Y., Zeng, J.-H., Lee, S.-Y., Lu, R.-B., and Kuo, P.-H., “SNP data science for classification of bipolar disorder I and bipolar disorder II,” IEEE/ACM Transactions on Computational Biology and Bioinformatics, 18(6), 2862-2869, 2021.
[59] Li, J., Cheng, K., Wang, S., Morstatter, F., Trevino, R. P., Tang, J., & Liu, H., “Feature selection: a data perspective,” ACM Computing Surveys, 50(6), 1-45, 2017.
[60] Ding, Y., Ding, P., Zhao, X., Cao, Y., & Jia, M., “Transfer learning for remaining useful life prediction across operating conditions based on multisource domain adaptation,” IEEE/ASME Transactions on Mechatronics, 27(5), 4143–4152, 2022
[61] Nectoux, P., Gouriveau, R., Medjaher, K., Ramasso, E., Chebel-Morello, B., Zerhouni, N., & Varnier, C., “PRONOSTIA: an experimental platform for bearings accelerated degradation tests,” 2012 IEEE International Conference on Prognostics and Health Management, 1-8. 2012.
[62] Wang, B., Lei, Y., Li, N., & Li, N., “A hybrid prognostics approach for estimating remaining useful life of rolling element bearings,” IEEE Transactions on Reliability, 69(1), 401-412, 2018.
[63] Zou, H., & Hastie, T., “Regularization and variable selection via the elastic net,” Journal of the Royal Statistical Society: Series B, 67(2), 301-320, 2005.
[64] van der Maaten, L., & Hinton, G. “Visualizing data using t-SNE,” Journal of Machine Learning Research, 9(86), 2579−2605, 2008.