Online remaining useful life (RUL) prediction has been solved by deep transfer learning, but is still with challenges as follows: (1) Incomplete and unlabeled data under actual operation; (2) Condition monitoring data is streaming with unknown distribution; and (3) The distribution of degradation data is variable. To solve them, an unsupervised incremental transfer learning approach with knowledge distillation (KD) is proposed. First, a time series recursive prediction model is built to generate pseudo labels. Second, an online KD network is constructed to realize unsupervised domain adaptation. Finally, an incremental updating mechanism is designed in the online KD network for online RUL prediction with the pseudo labels. Comparative experiments on the IEEE PHM Challenge 2012 rolling bearing dataset show that the proposed method, being computationally inexpensive, can effectively achieve dynamic prediction only with sequentially-collected data, which can provide an effective RUL prediction solution for rotating machinery under an open environment.
ChenJHLiWZhangWY, et al. Review of monitoring methods and technology of intelligent rolling bearing. Mech Strength2021; 43(3): 509–516.
2.
RaiAUpadhyaySH. Intelligent bearing performance degradation assessment and remaining useful life prediction based on self-organising map and support vector regression. J Mech Eng Sci2018; 232(6): 1118–1132.
3.
MahamadAKSaonSHiyamaT. Predicting remaining useful life of rotating machinery based artificial neural network. Comput Math Appl2010; 60(4): 1078–1087.
4.
PrasadAMIversonLRLiawA. Newer classification and regression tree techniques: bagging and random forests for ecological prediction. Ecosystems2006; 9(2): 181–199.
5.
GuoLLeiYGLiNP, et al. Machinery health indicator construction based on convolutional neural networks considering trend burr. Neurocomputing2018; 292: 142–150.
6.
MalhotraPVishnuTVRamakrishnanA, et al. Multi-sensor prognostics using an Unsupervised Health Index based on LSTM encoder-decoder. In: 1st ACM SIGKDD Workshop on Machine Learning for Prognostics and Health Management, San Francisco, CA, August 13–17, 2016. ACM: New York.
7.
CostaPROAkçayAZhangYQ, et al. Remaining useful lifetime prediction via deep domain adaptation. Reliabi Eng Syst Saf2020; 195: 106682.
8.
ZhangWLiXMaH, et al. Transfer learning using deep representation regularization in remaining useful life prediction across operating conditions. Reliabi Eng Syst Saf, 2021; 211: 107556.
9.
ZengFCLiYMJiangYH, et al. An online transfer learning-based remaining useful life prediction method of ball bearings. Measurement2021; 176: 109201.
10.
XuQChenZHWuKY, et al. KDnet-RUL: A knowledge distillation framework to compress deep neural networks for machine remaining useful life prediction. IEEE Trans Ind Electron2021; 69(2): 2022–2032.
11.
ZhouYQWangJWangZ. Bearing faulty prediction method based on federated transfer learning and knowledge distillation. Machines2022; 10(5): 376.
12.
RenLWangTJiaZD, et al. A lightweight and adaptive knowledge distillation framework for remaining useful life prediction. IEEE Transactions on Industrial Informatics, 2023; 19: 9060–9070. DOI: 10.1109/TII.2022.3224969.
13.
YuRZhengSAnandkumarA, et al. Long-term forecasting using higher order tensor RNNs. arXiv preprint, arXiv:1711.00073, 2017.
14.
MaoWTHeJLZuoMJ. Predicting remaining useful life of rolling bearings based on deep feature representation and transfer learning. IEEE Trans Instrum Meas2019; 69(4): 1594–1608.
15.
SunBCSaenkoK. 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, pp. 443–450. Springer International Publishing. Springer: Cham.
16.
YuXLiangZTWangYJ, et al. A wavelet packet transform-based deep feature transfer learning method for bearing fault diagnosis under different working conditions. Measurement2022; 201: 111597.
17.
MirzadehSIFarajtabarMLiA, et al. Improved knowledge distillation via teacher assistant. Proc AAAI Conf Artif Intell2020; 34(4): 5191–5198.
18.
ChenDFMeiJPWangC, et al. Online knowledge distillation with diverse peers, Proc AAAI Conf Artif Intell2020; 34(4): 3430–3437.
19.
NectouxPGouriveauRMedjaherK, et al. Pronostia: an experimental platform for bearings accelerated degradation test. In: IEEE international conference on prognostics and health management, PHM’12, Nanjing, Jiangsu, IEEE catalog number: CPF12PHMCDR, June 18–21, 2012, pp.1–8. IEEE: New York.
20.
MaoWTChenJXChenYJ, et al. Construction of health indicators for rotating machinery using deep transfer learning with multiscale feature representation. IEEE Trans Instrum Meas2021; 70: 1–13.
21.
DeutschJHeD. Using deep learning-based approach to predict remaining useful life of rotating components. IEEE Trans Syst Man Cybern Syst2017; 48(1): 11–20.
22.
MaoWTHeJLTangJM, et al. Predicting remaining useful life of rolling bearings based on deep feature representation and long short-term memory neural network. Adv Mech Eng2018; 10(12): 1–18.
23.
ZhuJChenNPengWW. Estimation of bearing remaining useful life based on multiscale convolutional neural network. IEEE Trans Ind Electron2018; 66(4): 3208–3216.
24.
HuangJYGrettonABorgwardtKM, et al. Correcting sample selection bias by unlabeled data. Advances in Neural Information Processing System 19: Proceedings of the 2006 Conference, MIT Press, 2007, pp. 601–608.
25.
FernandoBHabrardASebbanM, et al. Unsupervised visual domain adaptation using subspace alignment. In 2013 IEEE international conference on computer vision, Sydney, NSW, Australia, Dec 1–8, 2013, pp.2960–2967. IEEE: Washington DC.
26.
GongBShiYShaF, et al. Geodesic flow kernel for unsupervised domain adaptation. In: 2012 IEEE conference on computer vision and pattern recognition, Providence, RI, USA, June 16–21, 2012, pp.2066–2073. IEEE: Washington DC.
27.
PanSJTsangIWKwokJT, et al. Domain adaptation via transfer component analysis. IEEE Trans Neural Netw2010; 22(2): 199–210.
28.
ChenJXMaoWTLiuJ, et al. Online remaining useful life evaluation of rolling bearings under unknown operating conditions based on time series transfer recursive prediction. Control Decis2023; 38(1): 112–122.
29.
ChengHKongXGChenG, et al. Transferable convolutional neural network based remaining useful life prediction of bearing under multiple failure behaviors. Measurement2021 ; 168: 108286.
30.
GaninYUstinovaEAjakanH, et al. Domainadversarial training of neural networks. J Mach Learn Res2016; 17(1): 2096–2030.
