Journal of Jilin University(Engineering and Technology Edition) ›› 2022, Vol. 52 ›› Issue (1): 53-62.doi: 10.13229/j.cnki.jdxbgxb20200751

Previous Articles     Next Articles

State recognition in bearing temperature of high-speed train based on machine learning algorithms

Liang DUAN(),Chun-yuan SONG,Chao LIU,Wei WEI,Cheng-ji LYU   

  1. National Railway Vehicle Engineering R&D Center，CRRC Changchun Railway Vehicles Co. ，Ltd. ，Changchun 130002，China
  • Received:2020-07-10 Online:2022-01-01 Published:2022-01-14

RICH HTML

 2   

PDF (PC)

182

Abstract:

By studying a large number of historical data of high-speed train bearing temperature， the opinion， which the trend among related measuring positions of bearing temperature are similar， was used to identify the state of bearing temperature. Firstly the unsupervised learning algorithm One-class SVM was applied to abnormality detection. The result shows that the model could detect the abnormality well but not identify the state consistently. Then the whole process， which meant the bearing worked from normal to failure according to the relative temperature， was divided into four phases， the long， medium， short and urgent phases. The way， which describes the trend of bearing temperature through image including the information of relative temperature and the changing rate of temperature， was proposed. Finally， the deep learning algorithm CNN was used to identify the bearing temperature state consistently. The result shows that the CNN model could reliably identify the state of bearing temperature， and this could provide some support for predictive maintenance.

Key words: vehicle engineering, bearing temperature state, convolutional neural network, image classification, high-speed train, machine learning

CLC Number: 

  • U271

Fig.1

Measuring position of axle box bearing"

Fig.2

Temperature of related position of normal bearing"

Fig.3

Temperature of related position of abnormal bearing"

Fig.4

Disassemble of abnormal bearing"

Table 1

Rule in labelling temperature state"

距故障发生/天温度状态标签
0~3紧急（含失效）
4~10短
11~31中
32~inf长

Fig.5

Historical temperature of related position in training set including label"

Fig.6

Historical temperature of related position in testing set including label"

Fig.7

Result of One-class hyperplane in training set"

Fig.8

Result of One-class hyperplane in testing set"

Fig.9

Temperature of related position of normal bearing in time domain"

Fig.10

Temperature of related position of abnormal bearing in time domain"

Fig.11

Relative temperature and changing rate of abnormal bearing before failure"

Fig.12

Relative temperature and changing rate of normal bearing before failure"

Fig.13

Data sample of related position in time domain"

Fig.14

Image sample used for identifying bearing temperature state"

Table 2

Rule in defining temperature state"

关联测点之间相对温度/℃温度状态标签
0~15长
15~30中
30~45短
45~inf紧急

Fig.15

Temperature state of two abnormal bearing before failure"

Table 3

Network structure of CNN model"

序号层类型特征图数量特征图大小
1输入层1300×300×3
2卷积层13255×55×32
3池化层13227×27×32
4卷积层26427×27×64
5池化层26413×13×64
6卷积层36413×13×64
7卷积层425613×13×256
8卷积层512813×13×128
9池化层31286×6×128
10全连接层11 0241×1×1 024
11全连接层21 0241×1×1 024
12输出层11×1×4

Fig.16

Accuracy and cross entropy during model training"

Fig.17

Confusion matrix of result of testing set"

Fig.18

Accuracy of CNN model in testing set"

1 Olver A V. The mechanism of rolling contact fatigue: an update[J]. Journal of Engineering Tribology, 2005, 219(5):313-330.
2 Sadeghi F, Jalalahmadi B, Slack T S, et al. A review of rolling contact fatigue[J]. Journal of Tribology, 2009,131 (4):1-15.
3 Harvey T J, Wood R J, Powrie H E. Electrostatic wear monitoring of rolling element bearings[J]. Wear, 2007, 263(7-12):1492-1501.
4 Tarawneh C M, Sotelo L. Temperature profilesof railroad tapered roller bearings with defective inner and outer rings[C]∥Proceedings of the ASME Joint Rail Conference, Columbia, USA,2015:12-15.
5 Kim W, Seo J, Hong D. Infrared thermographic inspection of ball bearing; condition monitoring for defects under dynamic loading stages[C]∥The 18th World Conference on Nondestructive Testing, Durban, South Africa, 2012:16-20.
6 Kim D, Yun H, Yang S, et al. Fault diagnosis of ball bearings within rotational machines using the infrared thermography method[J]. Journal of the Korean Society for Nondestructive Testing, 2010, 30(6):558-563.
7 Sciascera C, Giangrande P, Papini L, et al. Analytical thermal model for fast stator winding temperature prediction[J]. IEEE Transactions on Industrial Electronics, 2017, 64(8): 6116-6126.
8 Madonna V, Walker A, Giangrande P, et al. Improved thermal management and analysis for stator end-windings of electrical machines[J]. IEEE Transactions on Industrial Electronics, 2019, 66(7): 5057-5069.
9 Pouly F, Changenet C, Ville F, et al. Investigations on the power losses and thermal behavior of rolling element bearings[J]. Journal of Engineering Tribology, 2010, 224(9): 925-933.
10 Neurouth A, Changenet C, Ville F, et al. Thermal modeling of a grease lubricated thrust ball bearing[J]. Journal of Engineering Tribology, 2014, 228(11): 1266-1275.
11 丁洪福, 王风涛, 景敏卿,等. 高速球轴承热稳定性研究[J]. 振动与冲击, 2017, 36(14):168-173.
Ding Hong-fu, Wang Feng-tao, Jing Min-qing, et al. The thermal stability of high-speed ball bearings[J]. Journal of Vibration and Shock, 2017, 36(14):168-173.
12 胡腾, 殷国富, 邓聪颍. 角接触球轴承热力学耦合模型与分析方法[J]. 四川大学学报:工程科学版, 2014, 46(4):189-198.
Hu Teng, Yin Guo-fu, Deng Cong-ying. Integrated thermo-mechanical model and analysis of angular contact ball bearing[J]. Advanced Engineering Sciences, 2014, 46(4):189-198.
13 何兴平, 耿远松, 郭志伟,等. 基于差分自回归移动平均模型的电气设备温度预测[J]. 自动化与仪器仪表, 2016(12):96-98.
He Xing-ping, Geng Yuan-song, Guo Zhi-wei, et al. The substation equipment temperature prediction based on time sequence ARIMA model[J]. Automation & Instrumentation, 2016(12):96-98.
14 邹永久, 张跃文, 孙培廷,等. 基于 RKGM-AR 模型的船舶主机排气温度预测[J]. 船舶工程, 2015,37(7):46-49, 58.
Zou Yong-jiu, Zhang Yue-wen, Sun Pei-ting, et al. Prediction of exhaust gas temperature of ship main engine based on RKGM-AR model[J]. Ship Engineering, 2015,37(7):46-49, 58.
15 孙建平, 朱雯. 基于小波 BP-时间序列的齿轮箱温度预警[J]. 电子测量与仪器学报, 2012, 26(3):197-201.
Sun Jian-ping, Zhu Wen. Gearbox temperature prewarning based on wavelets BP-time series method[J]. Journal of Electronic Measurement and Instrument, 2012, 26(3):197-201.
16 王新全, 孙培廷, 邹永久,等. 基于 GA-BP 模型的船舶柴油机排气温度趋势预测[J]. 大连海事大学学报, 2015, 41(3):73-76.
Wang Xin-quan, Sun Pei-ting, Zou Yong-jiu, et al. Prediction of marine main engine exhaust temperature changing trend based on GA-BP model[J]. Journal of Dalian Maritime University, 2015, 41(3):73-76.
17 梅启智. 系统可靠性工程基础[M].北京:科学出版社, 1992.
18 刘丽娟,陈果,郝腾飞. 基于流形学习与一类支持向量机的滚动轴承早期故障识别方法[J]. 中国机械工程, 2013, 24(5):628-633.
Liu Li-juan, Chen Guo, Hao Teng-fei. Incipient fault recognition of rolling bearings based on manifold learning and one-class SVM[J]. China Mechanical Engineering, 2013, 24(5):628-633.
19 Arlot S,Celisse A. A survey of cross-validation procedures for model selection[J]. Statistics Surveys,2010,4:40-79.
20 龚丁禧,曹长荣.基于卷积神经网络的植物叶片分类[J].计算机与现代化, 2014(4):12-15, 19.
Gong Ding-xi, Cao Chang-rong. Plant leaf classification based on cnn[J]. Computer and Modernization, 2014 (4):12-15, 19.
21 董峻妃,郑伯川,杨泽静.基于卷积神经网络的车牌字符识别[J].计算机应用,2017,37(7):2014-2018.
Dong Jun-fei, Zheng Bo-chuan, Yang Ze-jing. Character recognition of license plate based on convolution neural network[J]. Journal of Computer Applications, 2017, 37(7):2014-2018.
22 Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks[C]∥The Twenty-sixth Annual Conference on Neural Information Processing System, Lake Tahoe,USA,2012:1097-1105.
[1] Xing-jun HU,Jing-long ZHANG,Yu-fei LUO,Li XIN,Sheng LI,Jin-rui HU,Wei LAN. Influence investigation of cooling tube structure and airflow direction on thermal⁃hydraulic performance of air⁃cooled charge air cooler [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(6): 1933-1942.
[2] Jian-bin CHEN,Song-ze ZHOU,Feng-yong FEI,Yong-long CHEN,Guo-ping LING. Influences of interference fit and knurling connection type on press fitting failure of assembled camshaft [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(6): 1959-1966.
[3] Hui ZHONG,Heng KANG,Ying-da LYU,Zhen-jian LI,Hong LI,Ruo-chuan OUYANG. Image manipulation localization algorithm based on channel attention convolutional neural networks [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(5): 1838-1844.
[4] Yong LUO,Yong-heng WEI,Huan HUANG,Ren-jie XIAO,Lin REN,Huan-yu CUI. Starting control of P2.5 plug⁃in hybrid configuration dual⁃clutch based on driver's intention recognition [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(5): 1575-1582.
[5] Xiao-hua ZENG,Mei-jie SONG,Da-feng SONG,Yue WANG. Data processing method of bus driving cycle based on vehicular network information [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(5): 1692-1699.
[6] Chao MA,Yun-kai GAO,Zhe LIU,Yue-xing DUAN,Lin-li TIAN. Optimization of multi⁃material and beam cross⁃sectional shape and dimension of skeleton⁃type body [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(5): 1583-1592.
[7] Feng-chong LAN,Ji-wen LI,Ji-qing CHEN. DG-SLAM algorithm for dynamic scene compound deep learning and parallel computing [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(4): 1437-1446.
[8] Jian YANG,Qi XIA,Hai-chao ZHOU,Guo-lin WANG. Noise reduction mechanism of truck radial tire based on modified carcass string contour design [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(4): 1198-1203.
[9] Jiang-qi LONG,Jin-tao XIANG,Ping YU,Jun-cheng WANG. Linear disturbance observer suitable for sliding mode control of nonlinear active suspension [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(4): 1230-1240.
[10] De-xing WANG,Ruo-you WU,Hong-chun YUAN,Peng GONG,Yue WANG. Underwater image restoration based on multi-scale attention fusion and convolutional neural network [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(4): 1396-1404.
[11] Xin CHEN,Gui-shen YU,Biao ZHANG,Kai-xuan PAN,Li-fei YANG. Equivalent modeling of tensile-shear behavior for friction stir spot welding joints [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(4): 1190-1197.
[12] Guang-song LI,Wen-qing LI,Qing LI. Encrypted and compressed traffic classification based on random feature set [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(4): 1375-1386.
[13] Hou⁃jie LI,Fa⁃sheng WANG,Jian⁃jun HE,Yu ZHOU,Wei LI,Yu⁃xuan DOU. Pseudo sample regularization Faster R⁃CNN for traffic sign detection [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(4): 1251-1260.
[14] Xiao-long ZHU,Zhong XIE. Geospatial data extraction algorithm based on machine learning [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(3): 1011-1016.
[15] Jia-xu ZHANG,Xin-zhi WANG,Jian ZHAO,Zheng-tang SHI. Path planning and discrete sliding mode tracking control for high⁃speed lane changing collision avoidance of vehicle [J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(3): 1081-1090.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
Copyright © Journal of Jilin University(Engineering and Technology Edition), All Rights Reserved.
Tel: +86-431-85095297　Fax: +86-431-85094074　E-mail: xbgxb@jlu.edu.cn
Powered by Beijing Magtech Co. Ltd