SCIENTIA SINICA Informationis, Volume 50 , Issue 4 : 527-539(2020) https://doi.org/10.1360/SSI-2019-0232

Data-driven multimodal operation monitoring and fault diagnosis of high-speed train bearings

More info
  • ReceivedOct 18, 2019
  • AcceptedJan 24, 2020
  • PublishedApr 13, 2020


The multiple bearings of high-speed trains operate in a similar condition that correlates the bearings' temperatures dynamically. This allows for detecting bearing abnormalities and diagnosing the faults according to unexpected changes in the dynamic relations. Traditional single-modal modeling methods do not consider the different dynamic characteristics while the train is in different operating zones. This leads to difficulty with modeling the relationship using a single model, resulting in false alarms. To solve the above problems, the bearing temperature data collected by the bearing temperature monitoring system, the location information of the historical bearing fault alarms collected by the train control system, and the maintenance data are incorporated to identify the operation modes. In this paper, a novel data-driven multimodal bearing operation monitoring and fault diagnosis method is proposed for high-speed trains. First, to manage the outliers and missing values caused by train network communication disconnection and sensor faults, a preprocessing method, which combines linear-filling and a dynamic principal component pursuit, is proposed to reconstruct the contaminated data of the bearing temperatures. Second, the train location information-based operation mode identification method is proposed according to the fact that the clustering results of multiple bearing temperature correlations match well with the operating zone, extracted from the location information of historical faults. Third, a multimodal dynamic inner canonical correlation analysis (M-DiCCA) modeling method and the corresponding operation monitoring method for the train bearings are proposed. Subsequently, a dynamic time warping (DTW)-based fault diagnosis method is proposed. Finally, the application results by using the bearing temperature data collected from practical train operations demonstrate the effectiveness of the proposed methods.

Funded by





[1] Randall R B. Vibration-based condition monitoring: industrial, aerospace and automotive applications. Mechanisms Machine Science, 2011, 3: 431--477. Google Scholar

[2] Yi C, Lin J, Zhang W. Faults Diagnostics of Railway Axle Bearings Based on IMF's Confidence Index Algorithm for Ensemble EMD.. Sensors, 2015, 15: 10991-11011 CrossRef PubMed Google Scholar

[3] Henao H, Kia S H, Capolino G A. Torsional-Vibration Assessment and Gear-Fault Diagnosis in Railway Traction System. IEEE Trans Ind Electron, 2011, 58: 1707-1717 CrossRef Google Scholar

[4] Sha M Y, Liu L G. Summary of bearing fault diagnosis technology based on vibration signal. Bearing, 2015, 9: 59--63. Google Scholar

[5] Yi C, Liu L, Lin J, et al. Quantizing the health state of railway axle bearing via signal-to-noise ratio defined by EEMD and SVD. In: Proceedings of 2018 IEEE International Conference on Prognostics and Health Management (ICPHM), Seattle, 2018. 1--7. Google Scholar

[6] P?aczek M, Kokot G. Modelling and Laboratory Tests of the Temperature Influence on the Efficiency of the Energy Harvesting System Based on MFC Piezoelectric Transducers.. Sensors, 2019, 19: 1558-1748 CrossRef PubMed Google Scholar

[7] Li J, Li M, Zhang J. Rolling bearing fault diagnosis based on time-delayed feedback monostable stochastic resonance and adaptive minimum entropy deconvolution. J Sound Vib, 2017, 401: 139-151 CrossRef ADS Google Scholar

[8] Harkat M F, Mourot G, Ragot J. Sensor failure detection of air quality monitoring network. IFAC Proc Vol, 2000, 33: 529--534. Google Scholar

[9] Kourti T, Macgregor J F. Process analysis, monitoring and diagnosis, using multivariate projection methods. Chemometrics Intell Lab Syst, 1995, 28: 3--21. Google Scholar

[10] Kourti T, Macgregor J F. Recent developments in multivariate SPC methods for monitoring and diagnosing process and product performance. J Qual Tech, 1996, 4: 409--428. Google Scholar

[11] Wang X, Kruger U, Irwin G W. Nonlinear PCA With the Local Approach for Diesel Engine Fault Detection and Diagnosis. IEEE Trans Contr Syst Technol, 2008, 16: 122-129 CrossRef Google Scholar

[12] Hamadache M, Lee D. Improving signal-to-noise ratio (SNR) for inchoate fault detection based on principal component analysis (PCA). In: Proceedings of the 14th International Conference on Control, Automation and Systems (ICCAS 2014), 2014. Google Scholar

[13] Garcia-Alvarez D, Fuente M J, Sainz G I. Fault detection and isolation in transient states using principal component analysis. J Process Control, 2012, 22: 551-563 CrossRef Google Scholar

[14] Benaicha A, Mourot G, Benothman K. Determination of principal component analysis models for sensor fault detection and isolation. Int J Control Autom Syst, 2013, 11: 296-305 CrossRef Google Scholar

[15] MacGregor J F, Jaeckle C, Kiparissides C. Process monitoring and diagnosis by multiblock PLS methods. AIChE J, 1994, 40: 826-838 CrossRef Google Scholar

[16] Liu Q, Fang T, Dong Y N, et al. Dynamic modeling and reconstruction based fault detection and diagnosis of train bearings. Acta Autom Sin, 2019, 45: 2233--2241. Google Scholar

[17] Dong Y N, Qin S J. Dynamic-inner canonical correlation and causality analysis for high dimensional time series data. IFAC-PapersOnLine, 2018, 51: 476--481. Google Scholar

[18] Dong Y, Qin S J. Dynamic latent variable analytics for process operations and control. Comput Chem Eng, 2018, 114: 69-80 CrossRef Google Scholar

[19] Hartigan J A, Wong M A. A K-means clustering algorithm. Appl Stat, 2013, 28: 100--108. Google Scholar

[20] Karypis G, Eui-Hong Han G, Kumar V. Chameleon: hierarchical clustering using dynamic modeling. Computer, 1999, 32: 68-75 CrossRef Google Scholar

[21] Zhao S J, Zhang J, Xu Y M. Monitoring of Processes with Multiple Operating Modes through Multiple Principle Component Analysis Models. Ind Eng Chem Res, 2004, 43: 7025-7035 CrossRef Google Scholar

[22] Zhao C, Wang F, Lu N. Stage-based soft-transition multiple PCA modeling and on-line monitoring strategy for batch processes. J Process Control, 2007, 17: 728-741 CrossRef Google Scholar

[23] Yao Y, Gao F. Phase and transition based batch process modeling and online monitoring. J Process Control, 2009, 19: 816-826 CrossRef Google Scholar

[24] Pang X S. The methods of comparison of missing data processing. Stat Dec, 2010, 2010: 152--155. Google Scholar

[25] Deshpande A, Dong Y N, Li G, et al. US Patent, US20160179599A1, 2016-06-23. Google Scholar

[26] Ku W, Storer R H, Georgakis C. Disturbance detection and isolation by dynamic principal component analysis. Chemometrics Intelligent Laboratory Syst, 1995, 30: 179-196 CrossRef Google Scholar

[27] Lennart L. System Identification: Theory for the User. Upper Saddle River: PTR Prentice Hall, 1999. 9--11. Google Scholar

[28] Candès E J, Li X, Ma Y. Robust principal component analysis?. J ACM, 2011, 58: 1-37 CrossRef Google Scholar

[29] Berndt D J, Clifford J. Using dynamic time warping to find patterns in time series. Seattle, 1994. Google Scholar

[30] Dong Y, Qin S J. A novel dynamic PCA algorithm for dynamic data modeling and process monitoring. J Process Control, 2018, 67: 1-11 CrossRef Google Scholar

[31] Yue H, Qin S J. Reconstruction-based fault identification using a combined index. Industrial Engineering Chemistry Research, 2001, 40(20): 4403-4414 https://doi.org/10.1021/ie000141+. Google Scholar

[32] Li G, Qin S J, Chai T, et al. Multi-directional reconstruction based contributions for root-cause diagnosis of dynamic processes. In: Proceedings of American Control Conference, 2014. 3500--3505. Google Scholar

Copyright 2020  CHINA SCIENCE PUBLISHING & MEDIA LTD.  中国科技出版传媒股份有限公司  版权所有

京ICP备14028887号-23       京公网安备11010102003388号