Bearing fault diagnosis method based on MTF-CNN

PDF(1616 KB)

Journal of Vibration and Shock ›› 2023, Vol. 42 ›› Issue (2) : 126-131.

ZHAO Zhihong1,2，LI Chunxiu1，DOU Guangjian1，YANG Shaopu2

Author information +

History +

Abstract

Bearing fault diagnosis is very important to ensure the safety of mechanical equipment. In recent years, data-driven fault diagnosis methods have attracted the attention of researchers. Unlike the traditional fault feature extraction methods that rely on expert experience, the deep learning method can realize end-to-end automatic fault feature extraction and classification. In response to the problem that one-dimensional signal cannot fully exploit the relevant information between data when used as input to a convolutional neural network (CNN), a bearing fault diagnosis method based on a MTF-CNN is proposed. The collected vibration signals are encoded by the Markov transition field(MTF), the data correlation in different time intervals is obtained according to the transfer probability between data, and the corresponding feature map is generated. Then, it is input to a CNN to complete feature extraction and fault classification. The model is verified by the bearing dataset of Case Western Reserve University. The experimental results show that the fault diagnosis accuracy of the model is over 99.8%, and better generalization performance is obtained compared to other image coding methods.

Key words

fault diagnosis / deep learning / Markov transition field(MTF) / convolutional neural network(CNN)

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

ZHAO Zhihong1,2，LI Chunxiu1，DOU Guangjian1，YANG Shaopu2. Bearing fault diagnosis method based on MTF-CNN[J]. Journal of Vibration and Shock, 2023, 42(2): 126-131

References

[1] S. Khan, T. Yairi, A review on the application of deep
learning in system health management, Mechanical Systems & Signal Processing, 2018,107(1):241–265.
[2] B.P. Cai, H.L. Liu, M. Xie, A real-time fault diagnosis methodology of complex systems using object-oriented Bayesian networks, Mechanical Systems & Signal Processing. 2016 (80): 31- 44
[3] S.J. Dong, X.Y. Xu, R.X. Chen, Application of fuzzy C-means method and classification model of optimized k-nearest neighbor for fault diagnosis of bearing [J]. Journal of the Brazilian Society of Mechanical Sciences and Engineering, 2016 ,38(8):2255–2263.
[4] Konar, P., & Chattopadhyay, P. Bearing fault detection of induction motor using wavelet and Support Vector Machines (SVMs). Applied Soft Computing, 2011,11(6), 4203–4211.
[5] 张钰,陈珺,王晓峰,刘飞,周文晶,王志国.随机森林在滚动轴承故障诊断中的应用[J].计算机工程与应用,2018,54(06):100-104+114.
ZHANG Yu, CHEN Jun, WANG Xiaofeng, LIU Fei, ZHOU Wenjing, WANG Zhiguo.Application of random forest on rolling element bearings fault diagnosis[J].Computer Engineering and Applications,2018,54(06):100-104+114.
[6] 曲建岭,余路,袁涛,田沿平,高峰.基于一维卷积神经网络的滚动轴承自适应故障诊断算法[J].仪器仪表学报,2018,39(07):134-143.
Qu Jianling;Yu Lu;Yuan Tao;Tian Yanping;Gao Feng.Adaptive fault diagnosis algorithm for rolling bearings based on one-dimensional convolutional neural network[J].Chinese Journal of Scientific Instrument,2018,39(07):134-143.
[7] 赵敬娇,赵志宏,杨绍普.基于残差连接和1D-CNN的滚动轴承故障诊断研究[J].振动与冲击,2021,40(10):1-6.
ZHAO Jingjiao，ZHAO Zhihong，YANG Shaopu.Rolling bearing fault diagnosis based on residual connection and 1D-CNN[J].Journal of Vibration and Shock,2021,40(10):1-6.
[8] 张西宁,向宙,唐春华.一种深度卷积自编码网络及其在滚动轴承故障诊断中的应用[J].西安交通大学学报,2018,52(07):1-8+59.
ZHANG Xining;XIANG Zhou;TANG Chunhua.A Deep
Convolutional Auto-Encoding Neural Network and Its
Application in Bearing Fault Diagnosis[J].Journal Of Xi'an
Jiaotong University,2018,52(07):1-8+59.
[9] Wu Y，Yang F，Liu Y，et al.A Comparison of 1-D and 2-D Deep Convolutional Neural Networks in ECG Classification[R].IEEE Engineering in Medicine ang Biology Society，Hawaii，USA:IEEE，2018,324-327.
[10] WEN Long， LI Xinyu， GAO Liang， et al． A new
convolutional neural network-based data-driven fault diagnosis method［J］． IEEE Transactions on Industrial Electronics，2017( 11) : 2774777．
[11]李恒,张氢,秦仙蓉,孙远韬.基于短时傅里叶变换和卷积神经网络的轴承故障诊断方法[J].振动与冲击,2018,37(19):124-131.
LI Heng, ZHANG Qin, QIN Xianrong, SUN Yuantao.Fault diagnosis method for rolling bearings based on short-time Fourier transform and convolution neural network.Journal of Vibration and Shock，2018,37(19):124-131.
[12]袁建虎,韩涛,唐建,安立周.基于小波时频图和CNN的滚动轴承智能故障诊断方法[J].机械设计与研究,2017,33(02):93-97.
Yuan Jianhu, Han Tao, Tang Jian, an Lizhou. Intelligent fault diagnosis method of rolling bearing based on wavelet time frequency diagram and CNN [J]. Mechanical design and research, 2017,33 (02): 93-97.
[13] WANG Z，OATES T ．Imaging time-series to improve classification and imputation［C］∥ Proceedings of the 24th International Conference on Artificial Intelligence，2015．
[14]陈雯,吕王勇,李思奇,代娟,邓柙.两种马尔可夫链状态转移概率矩阵的估计与比较[J].重庆理工大学学报(自然科学),2021,35(08):217-223.
Chenwen,Lv Wangyong,Li Siqi,Daijuan,Dengxia.Estimation and Comparison of Two Markov Chain State Transition Probability Matrices.[J].Journal of Chongqing University of Technology ,2021,35(08):217-223.
[15] Zhao, J., Mao, X., & Chen, L. Speech emotion recognition using deep 1D & 2D CNN LSTM networks. Biomedical Signal Processing and Control, 2019,474(31): 312–323.
[16] 赵宇凯,徐高威,刘敏.基于VGG16迁移学习的轴承故障诊断方法[J].航天器环境工程,2020,37(05):446-451.
Zhao Yukai,，Xu Gaowei ，Liumin.Bearing fault diagnosis method based on vgg16 migration learning[J].Spacecraft Environment Engineering ,2020,37(05):446-451.
[17] KＲIZHEVSKY A，SUTSKEVEＲ I，HINTON G． Image
Net classification with deep convolutional neural networks［C］/ /NIPS． Curran Associates Inc，2012．
[18] Case Western Ｒeserve University Bearing Data Center［EB/OL］．https: / / cse groups． case． edu / bearingdata center / pages / download- data- file.2018.
[19] HOTH，AHN K K． Modeling and simulation of hydrostatic transmission system with energy regeneration using hydraulic accumulator ［J］． Journal of Mechanical Science and Technology，2010，24( 5) : 1163-1175．
[20]仝钰,庞新宇,魏子涵.基于GADF-CNN的滚动轴承故障诊断方法[J].振动与冲击,2021,40(05):247-253+260.
Tong Yu, Pang Xinyu, Wei Zihan. Fault diagnosis method of rolling bearing based on GADF-CNN [J]. Journal of Vibration and Shock, 2021,40 (05): 247-253 + 260.