基于逐次变分模态分解和CBAM-ResNet的滚动轴承故障诊断方法

PDF(1470 KB)

振动与冲击 ›› 2025, Vol. 44 ›› Issue (4) : 298-304.

故障诊断分析

基于逐次变分模态分解和CBAM-ResNet的滚动轴承故障诊断方法

陈志刚*1，2，陶子纯1，王衍学1，史梦瑶1

作者信息 +

Fault diagnosis method of rolling bearing based on SVMD and CBAM-ResNet

CHEN Zhigang*1,2， TAO Zichun1， WANG Yanxue1， SHI Mengyao1#br#

Author information +

文章历史 +

摘要

针对噪声背景下滚动轴承信号故障特征提取与智能诊断问题，本文提出基于逐次变分模态分解(successive variational mode decomposition, SVMD)以及注意力机制-残差神经网络(convolutional block attention module-residual neural network，CBAM-ResNet)的轴承故障诊断方法。首先对轴承振动信号进行SVMD分解成一系列本征模态分量，根据包络熵和峭度融合评价指标选择含故障特征明显的模态分量并重构；将重构信号进行短时傅里叶变换得到时频图像。之后利用CBAM能够自适应捕捉图形特征的特点，把重构信号的时频图像输入CBAM-ResNet模型进行特征提取和故障模式识别。在CBAM-ResNet模型训练过程中，使用迁移学习的方法初始化ResNet模型的参数来提高模型的泛化性。与其他传统模型相比，本文所提方法的分类准确率高达96.68%，具有更强的故障特征提取能力。实验结果表明，CBAM-ResNet模型在变工况环境下也具有较高的识别精度。

Abstract

To solve the problem of fault feature extraction and intelligent diagnosis of rolling bearing signals, this paper proposed a bearing fault diagnosis method, based on successive variational mode decomposition (SVMD) and convolutional block attention module-residual neural network (CBAM-ResNet). It entailed decomposing bearing vibration signals using SVMD into a series of intrinsic mode components. The selection of components with distinct fault features was determined based on envelope entropy and kurtosis fusion evaluation indicators, followed by a reconstruction process. The reconstructed signals underwent transformation into time-frequency images using Short-Time Fourier Transform. After that, CBAM was able to capture the features of the graphic features adaptively, and the time-frequency images of the reconstructed signal were input into CBAM-ResNet model for feature extraction and fault pattern recognition. In the process of CBAM-ResNet model training, transfer learning was used to initialize ResNet model parameters to improve the generalization of the model. Compared with other traditional models, the classification accuracy of the proposed model is as high as 96.68%, and it has stronger fault feature extraction ability. The experimental results show that CBAM-ResNet model also has high recognition accuracy under variable working conditions.

导出引用

陈志刚1, 2, 陶子纯1, 王衍学1, 史梦瑶1. 基于逐次变分模态分解和CBAM-ResNet的滚动轴承故障诊断方法[J]. 振动与冲击, 2025, 44(4): 298-304

CHEN Zhigang1, 2, TAO Zichun1, WANG Yanxue1, SHI Mengyao1. Fault diagnosis method of rolling bearing based on SVMD and CBAM-ResNet[J]. Journal of Vibration and Shock, 2025, 44(4): 298-304

参考文献

[1] 姜云龙. 基于增强微弱故障信息的滚动轴承故障诊断研究[D]. 北京建筑大学, 2023.
JIANG Y L. Research on Rolling Bearing Fault Diagnosis Based on Enhanced Weak Fault Information[D]. Beijing University of Civil Engineering and Architecture, 2023.
[2] 王锦鸿, 李鸿光, 张文笛, 等. 自适应模态总数变分模态分解方法及其性能评估[J]. 振动与冲击, 2023, 42(10): 251-262.
WANG J H, LI H G, ZHANG W D, et al. Adaptive modal total variational mode decomposition method and its performance evaluation[J]. Journal of Vibration and Shock, 2023, 42(10): 251-262.
[3] HUANG N E, SHEN Z, LONG S R, et al. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis [J]. Proceedings of the Royal Society of London Series A: Mathematical, Physical and Engineering Sciences, 1998, 454(1971): 903-995
[4] 李军宁, 罗文广, 陈武阁. 面向振动信号的滚动轴承故障诊断算法综述[J]. 西安工业大学学报, 2022, 42(02): 105-122.
LI J N, LOU W G, CHEN W G. Overview of Algorithms for Rolling Bearing Fault Diagnosis Based on Vibration Signal[J]. Journal of Xi'an Technological University, 2022, 42(02): 105-122.
[5] DRAGOMIRETSKIY K, ZOSSO D. Variational mode decomposition[J]. IEEE Transactions on Signal Processing, 2014, 62(3): 531-544.
[6] 何勇, 王红, 谷穗. 一种基于遗传算法的VMD参数优化轴承故障诊断新方法[J]. 振动与冲击, 2021, 40(06): 184-189.
HE Y, W H, G S. New fault diagnosis approach for bearings based on parameter optimized VMD and genetic algorithm[J]. Journal of Vibration and Shock, 2021, 40(06): 184-189.
[7] Nazari M, SAKHAEI S. Successive variational mode decomposition[J]. Signal Processing, 2020, 174: 107610-10．
[8] 方学宠, 娄益凡, 吴安定, 等. 基于SVMD与SLLE的机械设备齿轮箱故障诊断方法[J]. 机械与电子, 2022, 40(01): 36-41+47.
FANG X C, LOU Y F, WU A D, et al. Fault Diagnosis Method of Gearbox in Mechanical Equipment Based on SVMD and SLLE[J]. Machinery & Electronics, 2022, 40(01): 36-41+47.
[9] 杜小磊, 陈志刚, 王衍学, 等. 形态经验小波变换和改进分形网络在轴承故障识别中的应用[J]. 振动工程学报, 2021, 34(03): 654-662.
DU X L, CHEN Z G, WANG Y X, et al. Application of morphological empirical wavelet transform and IFractalNet in bearing fault identification[J]. Journal of Vibration Engineering, 2021, 34(03): 654-662.
[10] 周翔宇, 毛善君, 李梅. 基于频域降采样和CNN的轴承故障诊断方法[J]. 北京大学学报(自然科学版), 2023, 59(02): 251-260.
ZHOU X Y, MAO S J, LI M. Bearing Fault Diagnosis Method Based on Down-Sampling in Frequency Domain and CNN[J]. Acta Scientiarum Naturalium Universitatis Pekinensis, 2023, 59(02): 251-260.
[11] 于波, 李建成, 陈先瑞, 等. 基于CNN的轴承变工况故障识别系统[J]. 电子测量技术, 2022, 45(19): 25-29.
YU B, LI J C, CHEN X R, et al. Bearing fault Identification system based on CNN[J]. Electronic Measurement Technology, 2022, 45(19): 25-29.
[12] 赵杰, 陈志刚, 赵志川, 等. 基于同步提取变换和DRSN的滚动轴承故障诊断研究[J]. 重庆理工大学学报(自然科学), 2021, 35(1): 138-144．
ZHAO J, CHEN Z G, ZHAO Z C, et al. Fault Diagnosis of Rolling Bearing Based on SET and DRSN[J] Journal of Chongqing University of Technology(Natural Science), 2021, 35(1): 138-144．
[13] 张宸嘉, 朱磊, 俞璐. 卷积神经网络中的注意力机制综述[J]. 计算机工程与应用, 2021, 57(20): 64-72.
ZHANG C J, ZHU L, YU L. Review of Attention Mechan Neural Networks[J]. Computer Engineering and Applications, 2021, 57(20): 64-72.
[14] XI Y H, LI M T, Zhou F, et al. SE-Inception-ResNet Model With Focal Loss for Transmission Line Fault Classification Under Class Imbalance[J]. IEEE Transactions on Instrumentation and Measurement, 2024. 73: 1-17.
[15] 赵小强, 张亚洲. 利用改进卷积神经网络的滚动轴承变工况故障诊断方法[J]. 西安交通大学学报, 2021, 55(12): 108-118.
ZHAO X Q, ZHANG Y Z. Improved CNN-Based Fault Diagnosis Method for Rolling Bearings under Variable Working Conditions[J]. Journal of Xi’an Jiaotong University, 2021, 55(12): 108-118.
[16] LIU S S, Yu K P. Successive multivariate variational mode decomposition based on instantaneous linear mixing model[J]. Signal Processing, 2022, 190: 108311.
[17] 唐贵基, 王晓龙. 参数优化变分模态分解方法在滚动轴承早期故障诊断中的应用[J]. 西安交通大学学报, 2015, 49(5): 73-81.
TANG G J, WANG X L. Parameter Optimized Variational Mode Decomposition Method with Application to Incipient Fault Diagnosis of Rolling Bearing[J]. Journal of Xi’an Jiaotong University, 2015, 49(5): 73-81.
[18] 石文杰, 黄鑫, 温广瑞, 等. 基于DS-VMD及相关峭度的滚动轴承故障诊断[J]. 振动.测试与诊断, 2021, 41(01): 133-141+204.
SHI W J, HUANG X, WEN G R, et al. Rolling Bearing Fault Diagnosis Based on DS-VMD and Correlated Kurtosis[J]. Journal of Vibration, Measurement & Diagnosis. 2021, 41(01): 133-141+204.
[19] WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[J]. European Conference on Computer Vision, 2018, 11211: 3-19.
[20] Szegedy, Christian, Vincent Vanhoucke, et al. Rethinking the Inception Architecture for Computer Vision[J]. 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, 2818–2826.