Deep learning-based fault diagnostic network of high-speed train secondary suspension systems for immunity to track irregularities and wheel wear

Abstract

Fault detection and isolation of high-speed train suspension systems is of critical importance to guarantee train running safety. Firstly, the existing methods concerning fault detection or isolation of train suspension systems are briefly reviewed and divided into two categories, i.e., model-based and data-driven approaches. The advantages and disadvantages of these two categories of approaches are briefly summarized. Secondly, a 1D convolution network-based fault diagnostic method for high-speed train suspension systems is designed. To improve the robustness of the method, a Gaussian white noise strategy (GWN-strategy) for immunity to track irregularities and an edge sample training strategy (EST-strategy) for immunity to wheel wear are proposed. The whole network is called GWN-EST-1DCNN method. Thirdly, to show the performance of this method, a multibody dynamics simulation model of a high-speed train is built to generate the lateral acceleration of a bogie frame corresponding to different track irregularities, wheel profiles, and secondary suspension faults. The simulated signals are then inputted into the diagnostic network, and the results show the correctness and superiority of the GWN-EST-1DCNN method. Finally, the 1DCNN method is further validated using tracking data of a CRH3 train running on a high-speed railway line.

References

[1.] Iwnicki S. Handbook of railway vehicle dynamics 2006 Boca Raton, FL CRC Taylor & Francis

[2.] Bruni S, Goodall R, Mei TX, Tsunashima H. Control and monitoring for railway vehicle dynamics. Veh Syst Dyn 2007, 45 7–8 743-779

[3.] Ye Y, Zhang Y, Wang Q, Wang Z, Teng Z, Zhang H. Fault diagnosis of high-speed train suspension systems using multiscale permutation entropy and linear local tangent space alignment. Mech Syst Sig Process 2020, 138 106565

[4.] Li C, Luo S, Cole C, Spiryagin M. An overview: modern techniques for railway vehicle on-board health monitoring systems. Veh Syst Dyn 2017, 55 7 1045-1070

[5.] Lebel D, Soize C, Funfschilling C, Perrin G. High-speed train suspension health monitoring using computational dynamics and acceleration measurements. Veh Syst Dyn 2020, 58 6 1-22

[6.] Wei X, Lin S, Liu H (2012) Distributed fault detection observer for rail vehicle suspension systems. In: 24th chinese control and decision conference (CCDC), Taiyuan. doi: https://doi.org/10.1109/ccdc.2012.6244541

[7.] Xue P, Chai X-D, Zheng S-B. Research on vehicle diagnosis based on state-space method. Artif Intell Res 2015, 4 2 55-60

[8.] Wei XK, Jia LM, Liu H. A comparative study on fault detection methods of rail vehicle suspension systems based on acceleration measurements. Veh Syst Dyn 2013, 51 5 700-720

[9.] Liu F, Zhang H, He X, Zhao Y, Gu F, Ball AD. Correlation signal subset-based stochastic subspace identification for an online identification of railway vehicle suspension systems. Veh Syst Dyn 2020, 58 4 569-589

[10.] Li P, Goodall R (2004) Model based condition monitoring for railway vehicle systems, Control 2004, University of Bath, 2004, ID-508

[11.] Wei X, Liu H, Qin Y (2011) Fault isolation of rail vehicle suspension systems by using similarity measure. In: International conference on intelligent railway transportation. Beijing, China, pp 391–396

[12.] Jesussek M, Ellermann K. Fault detection and isolation for a nonlinear railway vehicle suspension with a hybrid extended kalman filter. Veh Syst Dyn 2013, 51 10 1489-1501

[13.] Jesussek M, Ellermann K. Fault detection and isolation for a full-scale railway vehicle suspension with multiple Kalman filters. Veh Syst Dyn 2014, 52 12 1695-1715

[14.] Jesussek M, Ellermann K. Fault detection and isolation for a railway vehicle by evaluating estimation residuals. Procedia IUTAM 2015, 13 14-23

[15.] Onat A, Kılınç O, Lata M. A linear kalman filtering scheme for estimation of secondary vertical suspension of railway vehicles. Vibroeng PROCEDIA 2016, 7 124-128

[16.] Zoljic-Beglerovic S, Stettinger G, Luber B, Horn M. Railway suspension system fault diagnosis using cubature kalman filter techniques. IFAC-Papers Online 2018, 51 24 1330-1335

[17.] Wei X, Liu H, Jia L (2012) Fault detection of urban rail vehicle suspension system based on acceleration measurements. In: 2012 IEEE/ASME international conference on advanced intelligent mechatronics (AIM), Kaohsiung. pp. 1129–1134

[18.] Xu B, Zhang J, Guan X. Estimation of the parameters of a railway vehicle suspension using model-based filters with uncertainties, proceedings of the institution of mechanical engineers. Part F J Rail Rapid Transit 2014, 229 7 785-797

[19.] Hayashi Y, Kojima T, Tsunashima H, Marumo Y (2006) Real time fault detection of railway vehicles and tracks. In: IET international conference on railway condition monitoring, Birmingham. pp. 20–25

[20.] Hayashi Y, Tsunashima H, Marumo Y (2006) Fault detection of railway vehicles using multiple model approach. In: 2006 SICE-ICASE international joint conference, Busan. pp. 2812–2817

[21.] Mori H, Tsunashima H (2010) Condition monitoring of railway vehicle suspension using multiple model approach. In: International conference on control, automation and systems, Gyeonggi-do. pp. 584–589

[22.] Li P, Goodall R, Kadirkamanathan V (2003) Parameter estimation of railway vehicle dynamic model using rao-blackwellised particle filter. In: 2003 European control conference (ECC), Cambridge. pp. 2384–2389

[23.] Li P, Kadirkamanathan V, Goodall R. Estimation of parameters in a linear state space model using a rao-blackwellised particle filter. IEE Proceed Control Theory Appl 2004, 151 6 727-738

[24.] Li P, Goodall R, Weston P, Ling CS, Goodman C, Roberts C. Estimation of railway vehicle suspension parameters for condition monitoring. Control Eng Pract 2007, 15 1 43-55

[25.] Liu X, Alfi S, Bruni S. An efficient recursive least square-based condition monitoring approach for a rail vehicle suspension system. Veh Syst Dyn 2016, 54 6 814-830

[26.] Liu X, Alfi S, Bruni S (2016) An efficient condition monitoring strategy of railway vehicle suspension based on recursive least-square algorithm. In: The dynamics of vehicles on roads and tracks, Boca Raton, CRC Press

[27.] Gasparetto L, Alfi S, Bruni S. Data-driven condition-based monitoring of high-speed railway bogies. Int J Rail Transp 2013, 1 1–2 42-56

[28.] Alfi S, Bruni S, Mazzola L, Bionda S (2011) Model-based fault detection in railway bogies. In: 22nd international symposium on dynamics of vehicles on roads and tracks, Manchester, UK

[29.] Mei T, Ding X. A model-less technique for the fault detection of rail vehicle suspensions. Veh Syst Dyn 2008, 46 sup1 277-287

[30.] Mei TX, Ding XJ. Condition monitoring of rail vehicle suspensions based on changes in system dynamic interactions. Veh Syst Dyn 2009, 47 9 1167-1181

[31.] Dumitriu M. Fault detection of damper in railway vehicle suspension based on the cross-correlation analysis of bogie accelerations. Mech Ind 2019, 20 1 1-14

[32.] Aravanis TCI, Sakellariou JS, Fassois SD (2017) Vibration based fault detection under variable non measurable operating conditions via a stochastic functional model method and application to railway vehicle suspensions. In: proceedings of the surveillance 9 international conference, Morocco

[33.] Aravanis TCI, Sakellariou JS, Fassois SD (2018) Railway suspension fault detection under variable operating conditions via random vibration signals and the stochastic functional model based method. In: international conference on noise and vibration engineering (ISMA2018), Leuven

[34.] Aravanis TCI, Sakellariou JS, Fassois SD (2018) On the problem of random vibration based fault detection in railway vehicle suspensions under variable and non-measurable operating conditions. In: AVT-305 research specialists’ meeting on sensing systems for integrated vehicle health management for military vehicles, Athens

[35.] Aravanis TCI, Sakellariou J, Fassois S. A stochastic functional model based method for random vibration based robust fault detection under variable non–measurable operating conditions with application to railway vehicle suspensions. J Sound Vib 2020, 466 115006

[36.] Sakellariou JS, Petsounis KA, Fassois SD (2002) On board fault detection and identification in railway vehicle suspensions via a functional model based method. In: Proc. international conference on noise and vibration engineering (ISMA), Leuven

[37.] Sakellariou JS, Petsounis KA, Fassois SD. Vibration based fault diagnosis for railway vehicle suspensions via a functional model based method: a feasibility study. J Mech Sci Technol 2015, 29 2 471-484

[38.] Rossouw CCE, Fröhling RD, Nel AL (2019) Fault diagnosis in railway vehicle suspensions. In: International heavy haul association conference (IHHA), Narvik

[39.] Qin N, Jin WD, Huang J, Jiang P, Li ZM (2013) High-speed train bogie fault signal analysis based on wavelet entropy feature. In: Proceedings of the international conference on advanced engineering materials and technology, Zhangjiajie. pp.2286–2289

[40.] Kulkarni R, Qazizadeh A, Berg M, Stichel S (2019) Fault detection and isolation (FDI) method for vehicle running instability from vehicle dynamics response using machine learning. In: 11th international conference on railway bogies and running gears (BOGIE’19), Budapest

[41.] Wei X, Jia L, Guo K, Wu S. On fault isolation for rail vehicle suspension systems. Veh Syst Dyn 2014, 52 6 847-873

[42.] Hong N, Li L, Yao W, Zhao Y, Yi C, Lin J . High-speed rail suspension system health monitoring using multi-location vibration data. IEEE Trans Intell Transp Syst 2019, 21 7 2943-2955

[43.] Wu Y, Jin W, Ren J, Sun Z. Fault diagnosis of high-speed train bogie based on synchrony group convolutions. Shock Vib Spec Issue 2019

[44.] Yabuno H, Takano H, Okamoto H. Stabilization control of hunting motion of railway vehicle wheelset using gyroscopic damper. J Vib Control 2008, 14 1–2 209-230

[45.] True H, Asmund R. The dynamics of a railway freight wagon wheelset with dry friction damping. Veh Syst Dyn 2003, 38 2 149-163

[46.] Bruni S, Vinolas J, Berg M, Polach O, Stichel S. Modelling of suspension components in a rail vehicle dynamics context. Veh Syst Dyn 2011, 49 7 1021-1072

[47.] Braghin F, Bruni S, Diana G. Experimental and numerical investigation on the derailment of a railway wheelset with solid axle. Veh Syst Dyn 2006, 44 4 305-325

[48.] Jia F, Lei YG, Lu N, Xing SB. Deep normalized convolutional neural network for imbalanced fault classification of machinery and its understanding via visualization. Mech Syst Signal Process 2018, 110 349-367

[49.] Ali JB, Fnaiech N, Saidi L, Chebel-Morello B, Fnaiech F. Application of empirical mode decomposition and artificial neural network for automatic bearing fault diagnosis based on vibration signals. Appl Acoust 2015, 89 16-27

[50.] Worden K, Staszewski WJ, Hensman JJ. Natural computing for mechanical systems research: a tutorial overview. Mech Syst Signal Process 2011, 25 1 4-111

[51.] Zhang L, Xiong GL, Liu HS, Zou HJ, Guo WZ. Bearing fault diagnosis using multi-scale entropy and adaptive neuro-fuzzy inference. Expert Syst Appl 2010, 37 8 6077-6085

[52.] Rosa AD, Alfi S, Bruni S. Estimation of lateral and cross alignment in a railway track based on vehicle dynamics measurements. Mech Syst Signal Process 2019, 116 606-623

[53.] Tao GQ, Du X, Zhang JH, Wen ZF, Jin XS, Cui DB. Development and validation of a model for predicting wheel wear in high-speed trains. J Zhejiang Univ Sci A 2017, 18 8 603-616

[54.] Cui D, Wang H, Li L, Jin X. Optimal design of wheel profiles for high-speed trains, proceedings of the institution of mechanical engineers. Part F J Rail Rapid Transit 2013, 229 3 248-261

[55.] Ye YG, Ning J. Small-amplitude hunting diagnosis method for high-speed trains based on the bogie frame’s lateral–longitudinal–vertical data fusion, independent mode function reconstruction and linear local tangent space alignment, proceedings of the institution of mechanical engineers. Part F J Rail Rapid Transit 2009, 233 10 1050-1067

[56.] Krizhevsky A, Sutskever I, Hinton GE. ImageNet classification with deep convolutional neural networks. Commun ACM 2017, 60 6 84-90

[57.] Devries PMR, Viégas F, Wattenberg M, Meade BJ. Deep learning of aftershock patterns following large earthquakes. Nature 2018, 560 632-634

[58.] Huang P, Wen C, Fu L, Peng Q, Tang Y. A deep learning approach for multi-attribute data: a study of train delay prediction in railway systems. Inf Sci 2020, 516 234-253

[59.] Chang X, Tang B, Tan Q, Deng L, Zhang F. One-dimensional fully decoupled networks for fault diagnosis of planetary gearboxes. Mech Syst Signal Process 2019, 141 106482

[60.] Ye Y, Huang P, Sun Y, Shi D. MBSNet: a deep learning model for multibody dynamics simulation and its application to a vehicle-track system. Mech Syst Signal Process 2021, 157 107716

[61.] Meymand SZ, Keylin A, Ahmadian M. A survey of wheel–rail contact models for rail vehicles. Veh Syst Dyn 2016, 54 3 386-428

[62.] Kalker JJ. A fast algorithm for the simplified theory of rolling contact. Veh Syst Dyn 1982, 11 1 1-13

[63.] Zhai W. Vehicle–track coupled dynamics models 2020 Singapore Springer

[64.] Chaar N, Berg M. Simulation of vehicle–track interaction with flexible wheelsets, moving track models and field tests. Veh Syst Dyn 2006, 44 sup1 921-931

[65.] Smith ASJ, Odolinsk K, Niac SH, Jönssonc PA, Stichel S, Iwnicki S, Wheat P (2016) Estimating the marginal cost of different vehicle types on rail infrastructure. Working papers in transport economics, https://ideas.repec.org/p/hhs/ctswps/2016_026.html

[66.] Bezin Y, Pålsson BA (2021) Multibody simulation benchmark for dynamic vehicle–track interaction in switches and crossings: modelling description and simulation tasks. Veh Syst Dyn. https://doi.org/10.1080/00423114.2021.1942079

[67.] Ning J, Lin J, Zhang B. Time–frequency processing of track irregularities in high-speed train. Mech Syst Signal Process 2016, 66–67 339-348

[68.] Xu L, Zhai W, Chen Z. On use of characteristic wavelengths of track irregularities to predict track portions with deteriorated wheel/rail forces. Mech Syst Signal Process 2018, 104 264-278

[69.] Xiao X, Sun Z, Shen W. A Kalman filter algorithm for identifying track irregularities of railway bridges using vehicle dynamic responses. Mech Syst Signal Process 2020, 138 106582

[70.] Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. https://arxiv.org/abs/1409.1556

[71.] Shi D, Ye Y, Gillwald M, Hecht M. Designing a lightweight 1D convolutional neural network with bayesian optimization for wheel flat detection using carbody accelerations. Int J Rail Transp 2020, 9 4 1-31

[72.] Nair V, and Hinton GE (2010) Rectified linear units improve restricted boltzmann machines. In: Proceedings of the 27th international conference on machine learning, Haifa. pp. 807–814

[73.] He K, Zhang X, Ren S, Sun J (2015) Delving deep into rectifiers: surpassing human-level performance on ImageNet classification. In: IEEE international conference on computer vision (ICCV), Santiago. pp. 1026–1034

[74.] Keras: Deep learning library for theano and tensorflow, https://www.datasciencecentral.com/profiles/blogs/keras-deep-learning-library-for-theano-and-tensorflow

[75.] Liu W, Wen Y, Yu Z, Li M, Raj B, Song L (2017) SphereFace: deep hypersphere embedding for face recognition. In: IEEE conference on computer vision and pattern recognition (CVPR), Honolulu. pp. 6738–6746

[76.] Wu ZH, Huang NE. Ensemble empirical mode decomposition: a noise assisted data analysis method. Adv Adapt Data Anal 2009, 1 1 1-41

[77.] Shi D, Sabanovic E, Rizzetto L, et al. (2021) Deep learning based virtual point tracking for real-time target-less dynamic displacement measurement in railway applications. https://arxiv.org/abs/2101.06702

[78.] Andrews DF. Plots of high-dimensional data. Biometrics 1972, 28 125-136

[79.] Moustafa RE. Andrews curves. Wiley Interdiscip Rev Comp Stat 2011, 3 4 373-382

Funding

National Natural Science Foundation of China(71871188); China Scholarship Council(201707000113); fundamental research funds for central universities of the central south university(2682021CX051)