A Diagnostic Approach for Rotating Machinery Using Multi-Scale Feature Attention Mechanism

被引:11
作者
Wu J. [1 ]
Ding E. [2 ,3 ]
Cui R. [1 ]
Liu J. [1 ]
机构
[1] Xuhai College, China University of Mining and Technology, Xuzhou, 221008, Jiangsu
[2] CUMT-IoT Perception Mine Research Center, China University of Mining and Technology, Xuzhou, 221008, Jiangsu
[3] School of Information and Control Engineering, China University of Mining and Technology, Xuzhou, 221008, Jiangsu
来源
Ding, Enjie | 1600年 / Xi'an Jiaotong University卷 / 54期
关键词
Attention mechanism; Fault diagnosis; Multi-scale feature extraction; Random dropout; Rotating machinery;
D O I
10.7652/xjtuxb202002007
中图分类号
学科分类号
摘要
A multi-scale end-to-end fault diagnosis method based on attention mechanism is proposed to deal with the problem that rotating machinery fault diagnosis needs complex feature extraction process and the diagnostic accuracy is low for noise-containing samples. A random dropout mechanism is introduced to suppress input noise at the input end. Taking advantage of the fact that fault signal has multiple modals, the vibration signals under different scales are obtained by using coarse-grained layer. Then multi-scale features are extracted by using fully convolutional networks and fused by using an attention mechanism. Finally, rotating machinery fault classification is realized based on a multi-classification function. The validity of the model is verified with the Case Western Reserve University bearing dataset and gearbox dataset, respectively. Experiments show that the fault recognition rate of the method is up to 100%. When the ratio of signal to noise is -4 dB, the fault recognition accuracies are 84.77% and 78.365% on the Case Western Reserve University bearing dataset F and on the gearbox dataset, respectively. The recognition accuracy is higher than that of other machine learning algorithms, implying the proposed method has strong anti-noise ability. © 2020, Editorial Office of Journal of Xi'an Jiaotong University. All right reserved.
引用
收藏
页码:51 / 58
页数:7
相关论文
共 19 条
  • [1] Zhang X., Luan Z., Liu X., Research review of rolling bearing fault diagnosis based on deep learning, Equipment Management and Maintenance, 18, pp. 130-133, (2017)
  • [2] Wang D., Tsui K.L., Statistical modeling of bearing degradation signals, IEEE Transactions on Reliability, 66, 4, pp. 1331-1344, (2017)
  • [3] Liu T., Chen J., Dong G., Rolling element bearing fault diagnosis based on frequency band entropy, Journal of Vibration and Shock, 33, 1, pp. 77-80, (2014)
  • [4] Deng F., Qiang Y., Yang S., Et al., A fault diagnosis method of rolling element bearings with adaptive frequency window empirical wavelet transform, Journal of Xi'an Jiaotong University, 52, 8, pp. 22-29, (2018)
  • [5] Zhang W., Li C., Peng G., Et al., A deep convolutional neural network with new training methods for bearing fault diagnosis under noisy environment and different working load, Mechanical Systems and Signal Processing, 100, pp. 439-453, (2018)
  • [6] Xiao X., Wang J., Zhang Y., Et al., A two-dimensional convolution neural network optimization method for bearing fault diagnosis, Proceedings of the CSEE, 39, 15, pp. 4558-4568, (2019)
  • [7] Qin Y., Wang X., Zou J., The optimized deep belief networks with improved logistic sigmoid units and their application in fault diagnosis for planetary gearboxes of wind turbines, IEEE Transactions on Industrial Electronics, 66, 5, pp. 3814-3824, (2019)
  • [8] Zhou F., Hu P., Yang S., Et al., A multimodal feature fusion-based deep learning method for online fault diagnosis of rotating machinery, Sensors, 18, 10, (2018)
  • [9] Yuan B., Wang C., Jiang F., Et al., Wavelet FCNN: a deep time series classification model for wind turbine blade icing detection
  • [10] Jiang G., He H., Yan J., Et al., Multiscale convolutional neural networks for fault diagnosis of wind turbine gearbox, IEEE Transactions on Industrial Electronics, 66, 4, pp. 3196-3207, (2019)