Self-supervised fast identification method of metro rail corrugation based on carriage inner noise

MA Chaozhi1, 2, WANG Yang2, ZHANG Shufang3, ZHONG Jie4, XIAO Hong2

Journal of Vibration and Shock ›› 2025, Vol. 44 ›› Issue (6) : 282-290.

PDF(2093 KB)
PDF(2093 KB)
Journal of Vibration and Shock ›› 2025, Vol. 44 ›› Issue (6) : 282-290.
TRANSPORTATION SCIENCE

Self-supervised fast identification method of metro rail corrugation based on carriage inner noise

  • MA Chaozhi*1,2, WANG Yang2, ZHANG Shufang3, ZHONG Jie4, XIAO Hong2
Author information +
History +

Abstract

Rapid identification of rail corrugation is essential for railway maintenance departments to plan maintenance activities. Using carriage inner noise data for quick identification of rail corrugation has significant advantages. To address the challenge of limited labeled carriage inner noise data and the difficulty of general learning from few labeled samples, a two-stage self-supervised rail corrugation recognition model based on "pretraining + fine-tuning" is proposed. First, a large amount of carriage inner noise data from operating subway trains was collected, and a time-frequency diagram dataset was created using wavelet denoising and wavelet transform techniques. During the pretraining stage, a denoising convolutional autoencoder was used to perform representation learning on numerous unlabeled time-frequency diagrams, and optimal pretraining parameters were obtained. In the fine-tuning stage, supervised learning was conducted on a small set of labeled time-frequency diagrams using an "encoder + classifier" approach. Finally, model experiments were carried out, and rail corrugation recognition was evaluated. The results show that the proposed model effectively extracts key features from the time-frequency diagrams. The detection accuracies for "no corrugation," "30-100 mm corrugation," "100-300 mm corrugation," and "30-100 mm + 100-300 mm mixed corrugation" are 95.62%, 96.46%, 92.38%, and 90.37%, respectively.

Key words

Urban rail transit / Rail corrugation / Carriage inner noise / Rapid detection / Self-supervised learning

Cite this article

Download Citations
MA Chaozhi1, 2, WANG Yang2, ZHANG Shufang3, ZHONG Jie4, XIAO Hong2.

Self-supervised fast identification method of metro rail corrugation based on carriage inner noise [J].

Journal of Vibration and Shock, 2025, 44(6): 282-290

References

[1] 王阳, 胡吉平, 肖宏, 等. 钢弹簧浮置板轨道振动特性及钢轨波磨成因分析[J]. 铁道科学与工程学报, 2023, 20(7): 2456-2466.
Wang Yang, Hu Jiping, Xiao Hong, et al. Vibration Characteristics of Spring-Steel Floating Slab Track and Cause Analysis of Rail Corrugation[J]. Journal of Railway Science and Engineering, 2023, 20(7): 2456-2466.
 [2] 周世祺, 崔晓璐, 杨冰. 高速铁路钢轨波磨高发区段弹条断裂机理研究[J]. 振动与冲击, 2023, 42(23): 267-275.
Zhou Shiqi, Cui Xiaolu, Yang Bing. Fracture Mechanism of Elastic Bar in High-Incidence Section of Rail Corrugation of High-Speed Railway[J]. Journal of Vibration and Shock, 2023, 42(23): 267-275.
 [3] 李伟,温泽峰,王衡禹,等.地铁钢轨波磨演化过程中的特性分析[J].机械工程学报,2018,54(04):70-78.
Li Wei, Wen Zefeng, Wang Hengyu, et al. Analysis on the Evolution Characteristics of Rail Corrugation on a Metro[J]. Journal of Mechanical Engineering, 2018,54(04):70-78.
[4] Ward C P, Weston P F, Stewart E J C, et al. Condition Monitoring Opportunities Using Vehicle-Based Sensors[J]. Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit, 2011, 225(2): 202-218.
 [5] Ng A K, Martua L, Sun G. Dynamic Modelling and Acceleration Signal Analysis of Rail Surface Defects for Enhanced Rail Condition Monitoring and Diagnosis[C]. 2019 4th International Conference on Intelligent Transportation Engineering: IEEE, 2019: 69-73.
 [6] Chia L, Bhardwaj B, Lu P, et al. Railroad Track Condition Monitoring Using Inertial Sensors and Digital Signal Processing: A Review[J]. IEEE Sensors Journal, 2019, 19(1): 25-33.
 [7] 晏兆晋, 高翠香, 徐晓迪, 等. 基于车辆响应的高速铁路周期性轨道短波病害时频特性分析[J]. 中国铁道科学, 2020, 41(1): 10-17.
Zhaojin Yan, Cuixiang Gao, Xiaodi X. U., et al. Time-Frequency Characteristics Analysis of Periodic Short Wave Disease of Track for High-Speed Railway Based On Vehicle Response[J]. China Railway Science, 2020, 41(1): 10-17.
 [8] Li S, Núñez A, Li Z, et al. Automatic Detection of Corrugation: Preliminary Results in the Dutch Network Using Axle Box Acceleration Measurements[C]. Proceedings of the 2015 Joint Rail Conference. San Jose, CA, USA.: ASME rail transportation division, 2015: 1-7.
 [9] Zhang Y, Li B, Li J, et al. Rail Corrugation Diagnosis Method Based On Parameter Optimization Vmd and Spwvd[C]. 2019 IEEE 16th International Conference on Networking, Sensing and Control. Banff, AB, Canada: IEEE, 2019: 110-115.
 [10] 王阳, 肖宏, 张智海, 等. 基于车内振动噪声响应的钢轨波磨状态评估方法[EB/OL].(2023-12-01) [2024-17-14].https://link.cnki.net/urlid/11.2187.th.20231129.1502.028.
[11] 刘金朝, 陈东生, 赵钢, 等. 评判高铁轨道短波不平顺病害的轨道冲击指数法[J]. 中国铁道科学, 2016, 37(4): 34-41.
Liu Jinzhao, Chen Dongsheng, Zhao Gang, et al. Track Impact Index Method for Evaluating Track Short Wave Irregularity of High Speed Railway[J]. China Railway Science, 2016, 37(4): 34-41.
[12] 刘金朝, 徐晓迪, 牛留斌, 等. 高速铁路轨道短波健康状态动态诊断模型和方法[J]. 铁道学报, 2021, 43(10): 69-74.
Liu Jinzhao, Xu Xiaodi, Niu Liubin, et al. Dynamic Inspection Model and Methodology for Track Short Wave Health Management for High-Speed Railway[J]. Journal of the China Railway Society, 2021, 43(10): 69-74.
[13] 李彦夫, 韩特. 基于深度学习的工业装备PHM研究综述[J]. 振动.测试与诊断, 2022, 42(05): 835-847.
Li Yanfu, Han Te. Deep Learning Based Industrial Equipment Prognostics and Health Management: a Review[J]. Journal of Vibration,Measurement & Diagnosis, 2022, 42(05): 835-847.
[14] 李可, 何坚光, 宿磊, 等. 基于CAE-GAN的滚动轴承故障诊断方法[J]. 振动与冲击, 2023, 42(23): 65-70.
Li Ke, He Jianguang, Su Lei, et al. Fault diagnosis method for rolling bearing based on CAE-GAN[J]. Journal of Vibration and Shock, 2023, 42(23): 65-70.
[15] 何庆, 利璐, 李晨钟, 等. 基于深度学习的轨道不平顺与车体垂向加速度映射模型[J]. 铁道学报, 2023: 1-11.
He Qing, Li Lu, Li Chenzhong, et al. A Mapping Model between Track Irregularity and Vertical Car⁃body Acceleration Based on Deep Learning[J]. Journal of the China Railway Society, 2023: 1-11.
[16] Zhuang L, Qi H, Zhang Z. The Automatic Rail Surface Multi-Flaw Identification Based On a Deep Learning Powered Framework[J]. IEEE Transactions On Intelligent Transportation Systems, 2022, 23(8): 12133-12143.
[17] Iliopoulos I A, Sakellariou J S, Fassois S D. Track Segment Automated Characterisation Via Railway – Vehicle – Based Random Vibration Signals and Statistical Time Series Methods[J]. Vehicle System Dynamics, 2022, 60(10): 3336-3357.
[18] Xie Q, Tao G, Lo S M, et al. A Data-Driven Convolutional Regression Scheme for On-Board and Quantitative Detection of Rail Corrugation Roughness[J]. Wear, 2023, 524-525: 204770.
[19] Li C, Sun H, Li W, et al. A Multitask Learning Method for Rail Corrugation Detection Using in-Vehicle Responses and Noise Data[J]. IEEE Transactions On Intelligent Transportation Systems, 2023: 1-14.
[20] Hinton G E, Zemel R. Autoencoders, minimum description length and Helmholtz free energy[J]. Advances in neural information processing systems, 1993, 6.
[21] He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 770-778.
[22] 冯青松, 周豪, 陈艳明, 等. 隧道内地铁列车车内噪声预测分析[J]. 振动与冲击, 2021, 40(15): 271-276.
Feng Qingsong, Zhou Hao, Chen Yanming, et al. Prediction and analysis of noise inside metro train in tunnel[J]. Journal of Vibration and Shock, 2021, 40(15): 271-276.
[23] 黄文杰, 贾小平, 周劲松, 等. 轮轨粗糙度对地铁车辆噪声影响试验分析[J]. 城市轨道交通研究, 2022, 25(12): 234-239.
Huang Wenjie, Jia Xiaoping, Zhou Jinsong, et al. Experimental Analysis of Wheel/Rail Roughness Influence on Metro Vehicle Noise[J]. Urban Mass Transit, 2022, 25(12): 234-239.
[24] 周月, 文永蓬, 王向阳, 等. 考虑滚动噪声的地铁波磨打磨限值评估方法[J]. 振动.测试与诊断, 2024, 44(01): 100-106.
Zhou Yue, Wen Yongpeng, Wang Xiangyang, et al. Evaluation Method of the Grinding Limit for Metro Rail Corrugation Considering the Rolling Noise[J]. Journal of Vibration,Measurement & Diagnosis, 2024, 44(01): 100-106.
Self-supervised fast identification method of metro rail corrugation based on carriage inner noise" title="Share on Weibo" target="_blank">
PDF(2093 KB)

241

Accesses

0

Citation

Detail

Sections
Recommended

/