volume | PIER Journals

Abstract

The problem of unstable vibration signal and accurate fault feature extraction of motor bearing fault causes the low accuracy of motor bearing fault diagnosis. In order to improve the accuracy of motor bearing fault diagnosis, the variational mode decomposition (VMD) is used to decompose the vibration signal and combine with the convolutional neural network (CNN).The bearing faults are categorized into inner ring wear, outer ring wear and cage fracture; then each category of faults is further subdivided into the degree of loading, which is categorized into 0, 25% and 50%, with a total of 9 cases. In order to select sensitive fault features, the vibration signals of motor bearings in three dimensions are collected, decomposed into multiple endowment modal function (IMF) components by VMD. The energy entropy of each IMF in each dimension is extracted, and the sensitive fault features are selected by feature selection (ReliefF), and then input into CNN for fault diagnosis. At the same time, the fault diagnosis of transverse vibration signal and three-dimensional vibration signal is also carried out respectively. The experimental results show that the accuracy of the method is greatly improved, and the fault diagnosis can be realized.

1. Shao, Siyu, Ruqiang Yan, Yadong Lu, Peng Wang, and Robert X. Gao, "DCNN-based multi-signal induction motor fault diagnosis," IEEE Transactions on Instrumentation and Measurement, Vol. 69, No. 6, 2658-2669, 2020. Google Scholar

2. Yang, Huixin, Xiang Li, and Wei Zhang, "Interpretability of deep convolutional neural networks on rolling bearing fault diagnosis," Measurement Science and Technology, Vol. 33, No. 5, 055005, 2022. Google Scholar

3. Feng, Donghua and Yahong Li, "Research on intelligent diagnosis method for large-scale ship engine fault in non-deterministic environment," Polish Maritime Research, Vol. 24, No. s3, 200-206, 2017.
doi:10.1515/pomr-2017-0123 Google Scholar

4. Wang, B., Design of motor condition detection and fault diagnosis system based on vibration characteristics, Qingdao University, 2021.

5. Xie, G. N., Y. Tong, and W. B. Lu, "Application of wavelet in fault diagnosis of coal mining machine asynchronous motor," Control Engineering, Vol. 20, No. 4, 2013. Google Scholar

6. Tao, Zan, Zhaoliang Pang, Min Wang, et al. "Early fault diagnosis method of rolling bearings based on VMD," Journal of Beijing Institute of Technology, Vol. 45, No. 2, 103-110, 2019. Google Scholar

7. Jin, Zhihao, Pengcheng Mu, Yimin Zhang, et al. "An improved VMD and its application in bearing fault diagnosis," Mechanical Design and Manufacturing, Vol. 2, 42-46, 2022. Google Scholar

8. Yuan, Laohu, Dongshan Lian, Xue Kang, Yuanqiang Chen, and Kejia Zhai, "Rolling bearing fault diagnosis based on convolutional neural network and support vector machine," IEEE Access, Vol. 8, 137395-137406, 2020. Google Scholar

9. Korobovaen, Sevalnevgs, Gromovvi, et al. "Steels for the manufacture of roller bearings for special purposes," Trudy VIAM, No. 11, 105, 2021. Google Scholar

10. Zhang, M., M. Yang, M. Yang, S. Li, et al. "Influence of carbon and nitrides in high-nitrogen stainless bearing steel on mechanical properties," Journal of Iron and Steel Research, Vol. 24, No. 5, 18-23, 2012. Google Scholar

11. Tang, G., L. Zhu, and X. Hu, "Rolling bearing fault diagnosis based on optimised VMD and deep confidence network," Bearing, No. 10, 47-53, 2020. Google Scholar

12. Zhao, G. Q., Z. D. Jiang, Cong Hu, Y. Gao, and G. Niu, "Bearing fault diagnosis based on wavelet packet energy entropy and DBN," Journal of Electronic Measurement and Instrumentation, Vol. 33, No. 2, 32-38, 2019. Google Scholar

13. Ince, Turker, Serkan Kiranyaz, Levent Eren, Murat Askar, and Moncef Gabbouj, "Real-time motor fault detection by 1-D convolutional neural networks," IEEE Transactions on Industrial Electronics, Vol. 63, No. 11, 7067-7075, Nov. 2016. Google Scholar

14. Tian, Shu and Zhiqi Kang, "Vibration analysis of circuit breaker mechanical failure based on improved variational modal decomposition and SVM," Vibration and Shock, Vol. 38, No. 23, 90-95, 2019. Google Scholar

15. Zhang, Lixin, Jiazhi Wang, Yannan Zhao, et al. "Relief-based combinatorial feature selection," Journal of Fudan (Natural Science Edition), No. 5, 893-898, 2004. Google Scholar

16. Li, Kunlun, Zefa Wei, and Huansheng Song, "Vehicle colour recognition based on SqueezeNet convolutional neural network," Journal of Chang'an University (Natural Science Edition), Vol. 40, No. 4, 109-116, 2020. Google Scholar

Mr. Ming Tang

Prof. Aiyuan Wang

Dr. Zhentian Zhu