融合CNN和ViT的声信号轴承故障诊断方法

PDF(2163 KB)

振动与冲击 ›› 2024, Vol. 43 ›› Issue (3) : 158-163.

论文

宁方立，王珂，郝明阳

作者信息 +

Fault diagnosis method for bearing based on fusing CNN and ViT

NING Fangli, WANG Ke, HAO Mingyang

Author information +

文章历史 +

摘要

针对轴承故障诊断任务数据量少、故障信号非平稳等特点，提出一种短时傅里叶变换、卷积神经网络和视觉转换器相结合的轴承故障诊断方法。首先，利用短时傅里叶变换将原始声信号转换为包含时序信息和频率信息的时频图像。其次，将时频图像作为卷积神经网络的输入，用于隐式提取图像的深层特征，其输出作为视觉转换器的输入。视觉转换器用于提取信号的时间序列信息。并在输出层利用Softmax函数实现故障模式的识别。实验结果表明，该方法对于轴承故障诊断准确率较高。为了更好解释和优化提出的轴承故障诊断方法，利用t-分布领域嵌入算法对分类特征进行了可视化展示。

Abstract

To the characteristics of small amount of data and non-stationary fault signal of bearing fault diagnosis, a bearing fault diagnosis method combining short-time Fourier transform, convolutional neural network and vision transformer is proposed. Firstly, the origin acoustic signal is converted by short-time Fourier transformer into a time-frequency image containing timing information and frequency information. Secondly, the time-frequency image is used as the input of the convolution neural network, which is used to implicitly extract the deep features of the image, and its output is the input of the vision transformer. And the vision transformer is used to extract time series information. Finally, the output layer is used to realize the pattern recognition of multiple faults of the bearing using Softmax function. The experimental results show that this method has a high accuracy rate for bearing fault diagnosis. In order to better explain and optimize the proposed bearing fault diagnosis method, the classification features are visualized by the t-distribution domain embedding algorithm.

导出引用

宁方立，王珂，郝明阳. 融合CNN和ViT的声信号轴承故障诊断方法[J]. 振动与冲击, 2024, 43(3): 158-163

NING Fangli, WANG Ke, HAO Mingyang. Fault diagnosis method for bearing based on fusing CNN and ViT[J]. Journal of Vibration and Shock, 2024, 43(3): 158-163

参考文献

[1] ALSHORMAN O, IRFAN M, SAAD N, et al. A review of artificial intelligence methods for condition monitoring and fault diagnosis of rolling element bearings for induction motor[J]. Shock and vibration, 2020,26:1-20. [2] 顾佶智,师蔚,胡定玉,等.强背景噪声下基于谱峭度-波束形成轴承故障特征提取[J].噪声与振动控制,2022,42(03):110-115. GU Yizhi, SHI Wei, HU Dingyu, et al. Characteristic extraction of bearing fault based on spectral steepness-beamforming under strong background noise[J].Noise and Vibration Control,2022,42(03):110-115. [3] CHEN, L; CHOY, YS; WANG T; CHIANG, YK, Fault detection of wheel in wheel/rail system using kurtosis beamforming method [J]. Structural Health Monitoring. 2020,(19):495-509. [4] CHEN, L; CHOY, YS; TAM, KC; FEI, CW, Hybrid microphone array signal processing approach for faulty wheel identification and ground impedance estimation in wheel-rail system [J]. Applied Acoustics. 2021,(172):10763 3. [5] 张秀峰,叶金山,黄苹.经验模态分解结合功率谱方法轴承故障诊断中的应用[J].机械工程师,2010,40(12):24-26. ZHANG Xiufeng, YE Jinshan, HUANG Ping. Application of empirical modal decomposition combined with power spectral method in bearing fault diagnosis[J].Mechanical Engineer, 2010,40(12):24-26 [6] 郑近德,程军圣.改进的希尔伯特-黄变换及其在滚动轴承故障诊断中的应用[J].机械工程学报,2015,51(01):138-145. ZHENG Jinde, CHENG Junsheng. Improved Hilbert-Yellow Transform and Its Application in Fault Diagnosis of Rolling Bearings[J].Chinese Journal of Mechanical Engineering,2015,51(01):138-145. [7] 王贡献,张淼,胡志辉，等.基于多尺度均值排列熵和参数优化支持向量机的轴承故障诊断[J].振动与冲击,2022,41(01):221-228. WANG Gongxian, ZHANG Miao, HU Zhihui, et al. Bearing fault diagnosis based on multi-scale mean arrangement entropy and parameter optimization support vector machine[J].Vibration and Shock,2022,41(01):221-22 8. [8] 路敦利,宁芊,杨晓敏.KNN-朴素贝叶斯算法的滚动轴承故障诊断[J].计算机测量与控制,2018,26(06):21-23+27. LU Dunli, NING Qian, YANG Xiaomin. Rolling bearing fault diagnosis of KNN-naïve Bayesian algorithm[J].Compu ter Measure mentand Control,2018,26(06):21-23+27. [9] 丁家满,吴晔辉,罗青波,等.基于深度森林的轴承故障诊断方法[J].振动与冲击,2021,40(12):107-113. DING Jiaman,WU Yehui,LUO Qingbo,et al. A fault diagnosis method of mechanical bearing based on the deep forest[J].Vibration and Shock,2021,40(12):107-113. [10] WANG Lihua, ZHAO Xiaoping, WU Jiaxin, et al. Motor fault diagnosis based on short-time Fourier transform and convolutional neural network[J]. Chinese Journal of Mechanical Engineering, 2017, 30(06): 1357-1368. [11] ZHU Zhiyu, PENG Gaoliang, CHEN Yuanhang, et al. A convolutional neural network based on a capsule network with strong generalization for bearing fault diagnosis[J]. Neurocomputing, 2019, 323: 62-75. [12] ZHANG Jiangquan, YI Sun, LIANG Guo, et al. A new bearing fault diagnosis method based on modified convolutional neural networks[J]. Chinese Journal of Aeronautics, 2020, 33(02): 439-447. [13] 吴楠,王喆.基于一维CNN与BiLSTM结合的轴承故障诊断[J].组合机床与自动化加工技术,2021,61(9):38-41+45. WU Nan, WANG Zhe. Bearing fault diagnosis based on the combination of one-dimensional CNN and BiLSTM[J]. Combined Machine Tool and Automated Processing Technology,2021,61(09):38-41+45. [14] CHEN Xiaohan, ZHANG Beike, GAO Dong. Bearing fault diagnosis base on multi-scale CNN and LSTM model[J]. Journal of Intelligent Manufacturing, 2021, 32(04): 971-987. [15] KHORRAM A, KHALOOEI M, REZGHI M. End-to-end CNN+ LSTM deep learning approach for bearing fault diagnosis[J]. Applied Intelligence, 2021, 51(02): 736-751. [16] SONG Ran, JIANG Quan. Application of VMD Combined with CNN and LSTM in Motor Bearing Fault[C]//2021 IEEE 16th Conference on Industrial Electronics and Applications (ICIEA). NJ :IEEE, 2021. 1661-1666. [17] MA Xiaolei, TAO Zhimin, WANG Yinhai.et al. Long short-term memory neural network for traffic speed prediction using remote microwave sensor data[J]. Transportation Research Part C:Emerging Technogies,2015, 54(01):187-197. [18] 陈子涵.基于多模态Transformer的机电作动器剩余寿命预测技术研究[EB/OL]. http://kns.cnki.net/kcms/detail/11.2 176.TJ.20221024.1556.003.html,2022-11-11. CHEN Zihan. Research on residual life prediction technology of electromechanical actuator based on multi-modal transformer [EB/OL].http://kns.cnki.net/kcms/ detail/11.2176.TJ.20221024.1556.003.html,2022-11-11. [19] VASWANI A, SHAZEER N, PARMARN, et al. Attention is all you need[C].NIPS'17:Proceedings of the 31st International Conference on Neural Information Processing Systems. New York, NY: Curran Associates,2017. 6000–6010 [20] DOSOVITSKY A, BEYER L, KOLESNIKOV A, et al. An image is worth 16x16 words: Transformers for image recognition at scale[J]. arXiv preprint arXiv:2010.11929,20 20. [21] Guo J, Han K, Wu H, et al. Cmt: Convolutional neural networks meet vision transformers[C]/ IEEE/CVF Conference on Computer Vision and Pattern Recognition..Piscataway, NJ: IEEE,2022.12175-12185. [22] NAKAMURA H, ASANO K, USUDA S, et al. A diagnosis method of bearing and stator fault in motor using rotating sound based on deep learning[J]. Energies, 2021, 14(05): 1319. [23] 赵海峰,杨国斌.基于振动信号的深沟球轴承滚动体故障诊断研究[J].化工机械,2014,41(03):291-295. ZHAO Haifeng, YANG Guobin. Research on fault diagnosis of rolling elements of deep groove ball bearings based on vibration signal[J].Chemical Machinery,2014,41(03):291-295