隨著科技和現代工業的飛速發展，工廠越來越依賴自動化生產設備。如果發生意外的設備故障，將會造成巨大的損失。為了防止生產設備出現故障，對其關鍵部件實施預防性維護策略是非常必要的。在預後和健康管理中，剩餘使用壽命預測可以幫助決策者安排維護任務和相關資源。本研究提出了一種混合方法，該方法結合了用於異常檢測的指數加權移動平均控製圖和深度學習模型，例如具有注意力模型的雙向長期短期記憶模型，並使用了貝葉斯優化尋找模型的超參數。本研究使用 IEEE PHM 2012 Data Challenge 數據集，使用 3 個數據集進行訓練，使用 11 個數據集進行測試。訓練數據從異常時間到生命結束的退化行為被用於具有注意力模型的雙向長期短期記憶模型中測試數據的遷移學習。評價標準 Score 的結果表明，所提出的方法優於其他五種現有方法。

With the rapid development of technology and modern industry, factories have become more dependent on automated production equipment. If an unexpected equipment failure occurs, it will cause huge losses. In order to prevent the failure of production equipment, it is extremely necessary to implement preventive maintenance strategies for its key parts. In prognosis and health management, the remaining useful life prediction can help decision makers to arrange maintenance tasks and related resources. This research proposes a hybrid method that combines an exponentially weighted moving average (EWMA) control chart for anomaly detection and a deep learning model, such as a bi-directional long short-term memory model with an attention mechanism (BiLSTM_Attention), and uses the Bayesian optimization to find the hyperparameters of the model. This study uses the IEEE PHM 2012 Data Challenge dataset, using 3 datasets for training and 11 datasets for testing. The degradation behavior of training data from the anomaly time to the end of life is used to transfer learning for the testing data in the BiLSTM_Attention model. The results of the evaluation criterion Score show that the proposed method is superior to the other five existing methods.

[1]Neupane D, Seok J. Bearing fault detection and diagnosis using case Western Reserve University dataset with deep learning approaches: a review. IEEE Access. 2020;8:93155-93178. https://doi.org/10.1109/ACCESS.2020.2990528
[2]Jardine A. K, Lin D, Banjevic D. A review on machinery diagnostics and prognostics implementing condition-based maintenance. Mech Syst Signal Process. 2006;20(7):1483-1510. https://doi.org/10.1016/j.ymssp.2005.09.012
[3]Cheng J. C, Chen W, Chen K, Wang Q. Data-driven predictive maintenance planning framework for MEP components based on BIM and IoT using machine learning algorithms. Autom. Constr. 2020;112:103087. https://doi.org/10.1016/j.autcon.2020.103087
[4]Lee J, Wu F, Zhao W, Ghaffari M, Liao L, Siegel D. Prognostics and health management design for rotary machinery systems—reviews, methodology and applications. Mech Syst Signal Process. 2014;42(1-2):314-334. https://doi.org/10.1016/j.ymssp.2013.06.004
[5]Yan R, Chen X, Li W, Sheng S. Mathematical methods and modeling in machine fault diagnosis. Math. Probl. Eng. 2014. https://doi.org/10.1155/2014/516590
[6]Wang D, Tsui K. L, Miao Q. Prognostics and health management: a review of vibration based bearing and gear health indicators. IEEE Access. 2018;6:665-676.
[7]Lei Y, Li N, Guo L, Li N, Yan T, Lin J. Machinery health prognostics: a systematic review from data acquisition to RUL prediction. Mech Syst Signal Process. 2018;104:799-834. https://doi.org/10.1016/j.ymssp.2017.11.016
[8]Zhao H, Liu H, Xu J, Deng W. Performance prediction using high-order differential mathematical morphology gradient spectrum entropy and extreme learning machine. IEEE Trans Instrum Meas. 2019;69(7):4165-4172. https://doi.org/10.1109/TIM.2019.2948414
[9]Dong S, Luo T. Bearing degradation process prediction based on the PCA and optimized LS-SVM model. Measurement, 2013;46(9):3143-3152. https://doi.org/10.1016/j.measurement.2013.06.038
[10]Sutrisno E, Oh H, Vasan AS, Pecht M. Estimation of remaining useful life of ball bearings using data driven methodologies. In: 2012 IEEE Conference on Prognostics and Health Management. Denver, CO, USA, 1‐7, 2012. https://doi.org/10.1109/ICPHM.2012.6299548
[11]Hong S, Zhou Z, Zio E, Hong K. Condition assessment for the performance degradation of bearing based on a combinatorial feature extraction method. Digit Signal Process. 2014;27(1):159-166. http://dx.doi.org/10.1016/j.dsp.2013.12.010
[12]Lei Y, Li N, Gontarz S, Lin J, Radkowski S, Dybala J. A model‐based method for remaining useful life prediction of machinery. IEEE Trans Reliab. 2016;65(3):1314‐1326. https://doi.org/10.1109/TR.2016.2570568
[13]Ren L, Lv W. Remaining useful life estimation of rolling bearings based on sparse representation. In: 2016 7th International Conference on Mechanical and Aerospace Engineering (ICMAE). London, UK, 209-213, 2016. https://doi.org/10.1109/ICMAE.2016.7549536
[14]Guo L, Li N, Jia F, Lei Y, Lin J. A recurrent neural network based health indicator for remaining useful life prediction of bearings. Neurocomputing. 2017;240:98‐109. http://dx.doi.org/10.1016/j.neucom.2017.02.045
[15]Liu X, Song P, Yang C, Hao C, Peng W. Prognostics and health management of bearings based on logarithmic linear recursive least-squares and recursive maximum likelihood estimation. IEEE Trans Ind Electron. 2017;65(2):1549-1558. https://doi.org/10.1109/TIE.2017.2733469
[16]Yoo Y, Baek J G. A novel image feature for the remaining useful lifetime prediction of bearings based on continuous wavelet transform and convolutional neural network. Appl Sci. 2018;8(7):1102. https://doi.org/10.3390/app8071102
[17]Zheng Y. Predicting remaining useful life based on Hilbert–Huang entropy with degradation model. Int J Electr Comput Eng. 2019;2019(99):1-11. https://doi.org/10.1155/2019/3203959
[18]Wang F K, Mamo T. Hybrid approach for remaining useful life prediction of ball bearings. Qual Reliab Eng Int. 2019;35(7):2494-2505. https://doi.org/10.1002/qre.238
[19]Chen Y, Peng G, Zhu Z, Li S. A novel deep learning method based on attention mechanism for bearing remaining useful life prediction. Appl Soft Comput. 2020; 86:105919. https://doi.org/10.1016/j.asoc.2019.105919
[20]Yang J, Yin S, Chang Y, Gao T, Yang C. An efficient method for monitoring degradation and predicting the remaining useful life of mechanical rotating components. IEEE Trans Instrum. 2020; 70:1-14. https://doi.org/10.1109/TIM.2020.3032218
[21]Yang H, Sun Z, Jiang G, Zhao F, Mei X. Remaining useful life prediction for machinery by establishing scaled-corrected health indicators. Measurement. 2020;163:108035. https://doi.org/10.1016/j.measurement.2020.108035
[22]Montgomery DC. Introduction to Statistical Quality Control. 7th ed. New York: Wiley; 2012.
[23]Zaremba W, Sutskever I, Vinyals O. Recurrent neural network regularization. v5. 2015. https://arxiv.org/abs/1409.2329
[24]Chorowski J, Bahdanau D, Serdyuk D, Cho K, Bengio Y. Attention-Based Models for Speech Recognition. 2015. https://arxiv.org/abs/1506.07503
[25]Nectoux P, Gouriveau R, Medjaher K, Ramasso E, Chebel-Morello B, Zerhouni N, Varnier C. PRONOSTIA: An experimental platform for bearings accelerated degradation tests. In: 2012 IEEE Conference on Prognostics and Health Management. Denver. CO, USA, 1-8, 2012.
[26]Mockus J, Mockus L. Bayesian approach to global optimization and application to multiobjective and constrained problems. J Optim Theory Appl. 1991; 70(1):157–172. https://doi.org/10.1007/BF00940509
[27]Sheu S H, Tai S H. Generally weighted moving average control chart for monitoring process variability. Int. J. Adv. Manuf. Technol. 2006; 60: 452-458. https://doi.org/10.1007/s00170-005-0091-0