留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

结合SE-VAE与M1DCNN的小样本数据下轴承故障诊断

李梦男 李琨 叶震 高宏宇

李梦男,李琨,叶震, 等. 结合SE-VAE与M1DCNN的小样本数据下轴承故障诊断[J]. 机械科学与技术,2024,43(5):773-780 doi: 10.13433/j.cnki.1003-8728.20230036
引用本文: 李梦男,李琨,叶震, 等. 结合SE-VAE与M1DCNN的小样本数据下轴承故障诊断[J]. 机械科学与技术,2024,43(5):773-780 doi: 10.13433/j.cnki.1003-8728.20230036
LI Mengnan, LI Kun, YE Zhen, GAO Hongyu. Bearing Fault Diagnosis Under Small Sample Data Based on SE-VAE and M1DCNN[J]. Mechanical Science and Technology for Aerospace Engineering, 2024, 43(5): 773-780. doi: 10.13433/j.cnki.1003-8728.20230036
Citation: LI Mengnan, LI Kun, YE Zhen, GAO Hongyu. Bearing Fault Diagnosis Under Small Sample Data Based on SE-VAE and M1DCNN[J]. Mechanical Science and Technology for Aerospace Engineering, 2024, 43(5): 773-780. doi: 10.13433/j.cnki.1003-8728.20230036

结合SE-VAE与M1DCNN的小样本数据下轴承故障诊断

doi: 10.13433/j.cnki.1003-8728.20230036
基金项目: 国家自然科学基金项目(82160787)与昆明理工大学科技园有限公司下达项目(2018KF3)
详细信息
    作者简介:

    李梦男,硕士研究生,lmncom@163.com

    通讯作者:

    李琨,副教授,博士,ghfighter@163.com

  • 中图分类号: TP277

Bearing Fault Diagnosis Under Small Sample Data Based on SE-VAE and M1DCNN

  • 摘要: 针对轴承故障诊断中故障样本数量少导致诊断正确率低的问题,提出了一种基于注意力机制变分自编码器(SE-VAE)和多尺度一维卷积神经网络(M1DCNN)的轴承故障诊断方法。将轴承数据集的训练集输入到SE-VAE中进行训练,生成与训练样本分布相似的生成样本,并添加到训练集中增加训练集的样本数量。将扩充后的训练集输入到M1DCNN中进行训练,随后将训练好的模型应用于测试集,输出故障诊断结果。实验结果表明,所提方法能够在不同负载的小样本轴承故障数据集上取得较好的故障诊断准确率。
  • 图  1  SENet网络结构图

    Figure  1.  SENet Network structure diagram

    图  2  VAE结构图

    Figure  2.  VAE Structure diagram

    图  3  原始Inception结构

    Figure  3.  Original Inception structure

    图  4  基于SE-VAE和M1DCNN的轴承故障诊断流程图

    Figure  4.  Flow chart of bearing fault diagnosis based on SE-VAE and M1DCNN

    图  5  SE-VAE模型结构图

    Figure  5.  Structure of SE-VAE model

    图  6  M1DCNN模型结构图

    Figure  6.  Structure of M1DCNN model

    图  7  数据采集系统

    Figure  7.  Data acquisition system

    图  8  SE-VAE模型Loss变化曲线

    Figure  8.  Loss variation curve of SE-VAE model

    图  9  标签为1,4,7的原始数据和生成数据的时频信号

    Figure  9.  The time-domain and frequency-domain signals of the raw and generated data corresponding to labels 1, 4, and 7

    图  10  M1DCNN训练集和验证集准确率变化曲线

    Figure  10.  Accuracy variation curves of M1DCNN training set and validation set

    表  1  实验数据集划分

    Table  1.   Experimental dataset division

    标签故障类型训练集测试集验证集
    0正常1005050
    10.177 8 mm内圈故障1005050
    20.355 6 mm内圈故障1005050
    30.533 4 mm内圈故障1005050
    40.177 8 mm滚动体故障1005050
    50.355 6 mm滚动体故障1005050
    60.533 4 mm滚动体故障1005050
    70.177 8 mm外圈故障1005050
    80.355 6 mm外圈故障1005050
    90.533 4 mm外圈故障1005050
    下载: 导出CSV

    表  2  添加不同数量生成样本对M1DCNN的影响

    Table  2.   The impact of incorporating varying quantities of generated samples on M1DCNN

    添加生成样本数量 平均准确率/% 准确率标准差
    0 96.38 0.015 56
    50 97.56 0.008 14
    100 98.11 0.007 78
    150 98.52 0.007 11
    200 98.84 0.005 28
    下载: 导出CSV

    表  3  不同负载下不同生成算法实验结果

    Table  3.   Experimental results of different generative algorithms under various loads

    生成数据
    算法
    负载
    大小/hp
    生成样本
    数量
    平均准
    确率/%
    准确率
    标准差
    1096.380.015 56
    SE-VAE110098.110.007 78
    VAE110097.920.010 85
    GAN110097.520.014 76
    SMOTE110045.120.031 57
    3095.570.013 77
    SE-VAE310097.870.009 65
    VAE310096.720.013 45
    GAN310096.620.017 55
    SMOTE310049.930.037 54
    下载: 导出CSV

    表  4  不同负载下不同诊断算法实验结果

    Table  4.   Experimental results of different diagnostic algorithms under different loads

    诊断算法负载大小/hp平均准确率/%准确率标准差
    M1DCNN196.380.015 56
    文献[19]195.000.021 91
    SVM189.370.024 00
    M1DCNN395.570.013 77
    文献[19]396.100.015 13
    SVM386.290.029 94
    下载: 导出CSV
  • [1] 牛一捷, 李花, 邓武, 等. 基于TQWT和稀疏表示的滚动轴承故障诊断方法[J]. 交通运输工程学报, 2021, 21(6): 237-246.

    NIU Y J, LI H, DENG W, et al. Rolling bearing fault diagnosis method based on TQWT and sparse representation[J]. Journal of Traffic and Transportation Engineering, 2021, 21(6): 237-246. (in Chinese)
    [2] 张伟. 基于卷积神经网络的轴承故障诊断算法研究[D]. 哈尔滨: 哈尔滨工业大学, 2017.

    ZHANG W. Study on bearing fault diagnosis algorithm based on convolutional neural network[D]. Harbin: Harbin Institute of Technology, 2017. (in Chinese)
    [3] HOANG D T, KANG H J. A survey on Deep Learning based bearing fault diagnosis[J]. Neurocomputing, 2019, 335: 327-335. doi: 10.1016/j.neucom.2018.06.078
    [4] FUAN W, HONGKAI J, HAIDONG S, et al. An adaptive deep convolutional neural network for rolling bearing fault diagnosis[J]. Measurement Science and Technology, 2017, 28(9): 095005. doi: 10.1088/1361-6501/aa6e22
    [5] HUANG W Y, CHENG J S, YANG Y, et al. An improved deep convolutional neural network with multi-scale information for bearing fault diagnosis[J]. Neurocomputing, 2019, 359: 77-92. doi: 10.1016/j.neucom.2019.05.052
    [6] 张笑璐, 邹益胜, 曾大懿, 等. 样本不均衡下的DCGAN轴承故障诊断方法[J]. 机械科学与技术, 2022, 41(1): 9-15.

    ZHANG X L, ZOU Y S, ZENG D Y, et al. DCGAN bearing fault diagnosis method under unbalanced samples[J]. Mechanical Science and Technology for Aerospace Engineering, 2022, 41(1): 9-15. (in Chinese)
    [7] WU Z H, JIANG H K, ZHAO K, et al. An adaptive deep transfer learning method for bearing fault diagnosis[J]. Measurement, 2020, 151: 107227. doi: 10.1016/j.measurement.2019.107227
    [8] 何强, 唐向红, 李传江, 等. 负载不平衡下小样本数据的轴承故障诊断[J]. 中国机械工程, 2021, 32(10): 1164-1171. doi: 10.3969/j.issn.1004-132X.2021.10.004

    HE Q, TANG X H, LI C J, et al. Bearing fault diagnosis method based on small sample data under unbalanced loads[J]. China Mechanical Engineering, 2021, 32(10): 1164-1171. (in Chinese) doi: 10.3969/j.issn.1004-132X.2021.10.004
    [9] WANG Z R, WANG J, WANG Y R. An intelligent diagnosis scheme based on generative adversarial learning deep neural networks and its application to planetary gearbox fault pattern recognition[J]. Neurocomputing, 2018, 310: 213-222. doi: 10.1016/j.neucom.2018.05.024
    [10] ZHAO D F, LIU S L, GU D, et al. Enhanced data-driven fault diagnosis for machines with small and unbalanced data based on variational auto-encoder[J]. Measurement Science and Technology, 2020, 31(3): 035004. doi: 10.1088/1361-6501/ab55f8
    [11] GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[C]// Proceedings of the 27th International Conference on Neural Information Processing Systems. Montreal: MIT Press, 2014: 2672-2680.
    [12] 翟正利, 梁振明, 周炜, 等. 变分自编码器模型综述[J]. 计算机工程与应用, 2019, 55(3): 1-9.

    ZHAI Z L, LIANG Z M, ZHOU W, et al. Research overview of variational auto-encoders models[J]. Computer Engineering and Applications, 2019, 55(3): 1-9. (in Chinese)
    [13] HU J, SHEN L, SUN G. Squeeze-and-excitation networks[C]// Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 7132-7141.
    [14] 任欢, 王旭光. 注意力机制综述[J]. 计算机应用, 2021, 41(增刊1): 1-6.

    REN H, WANG X G. Review of attention mechanism[J]. Journal of Computer Applications, 2021, 41(Sup.1): 1-6. (in Chinese)
    [15] KINGMA D P, WELLING M. Auto-encoding variational Bayes[C]// 2nd International Conference on Learning Representations. [S.I.]: Banff, 2014.
    [16] 胡铭菲, 左信, 刘建伟. 深度生成模型综述[J]. 自动化学报, 2022, 48(1): 40-74.

    HU M F, ZUO X, LIU J W. Survey on deep generative model[J]. Acta Automatica Sinica, 2022, 48(1): 40-74. (in Chinese)
    [17] SZEGEDY C, VANHOUCKE V, IOFFE S, et al. Rethinking the inception architecture for computer vision[C]// Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 2818-2826.
    [18] SMITH W A, RANDALL R B. Rolling element bearing diagnostics using the Case Western Reserve University data: a benchmark study[J]. Mechanical Systems and Signal Processing, 2015, 64-65: 100-131. doi: 10.1016/j.ymssp.2015.04.021
    [19] 张成帆, 江泽鹏, 曹伟, 等. 一维多尺度卷积神经网络及其在滚动轴承故障诊断中的应用[J]. 机械科学与技术, 2022, 41(1): 120-126.

    ZHANG C F, JIANG Z P, CAO W, et al. One-dimensional multi-scale convolution neural network and its application in rolling bearing fault diagnosis[J]. Mechanical Science and Technology for Aerospace Engineering, 2022, 41(1): 120-126. (in Chinese)
  • 加载中
图(10) / 表(4)
计量
  • 文章访问数:  49
  • HTML全文浏览量:  23
  • PDF下载量:  26
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-05-05
  • 刊出日期:  2024-05-31

目录

    /

    返回文章
    返回