基于子域自适应对抗网络的轴承故障诊断

摘要
图/表
参考文献(25)
相关文章 (15)

全文: PDF (3714 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要现有基于深度学习网络模型的故障诊断方法往往依赖大量有标签数据进行训练，在变工况条件下，模型的诊断精度会有所下降。针对此，为提高变工况条件下的故障诊断准确率，基于域自适应理论提出一种新颖的网络模型——子域自适应对抗网络。该网络模型不仅充分利用了动态卷积的特征提取能力，同时还借鉴了生成对抗网络的博弈思想，使特征生成器和分类器对抗学习，利用每个类别的决策边界对样本进行正确分类；此外，在对抗网络中引入局部最大平均差异，考虑每个类别的细粒度信息，以此来对齐源域和目标域相应的类空间，减小网络模型在决策边界附近的分类误差，从而提高模型对故障类别的识别精度。最终，通过两个数据集对所提出的方法进行实验验证，结果表明模型在变工况条件下具有较强的泛化性能与良好的故障识别精度。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	周华锋
	程培源
	邵思羽
	赵玉伟

关键词 ：故障诊断, 子域自适应, 动态卷积, 变工况

Abstract：Existing deep learning-based fault diagnosis methods mainly rely on sufficient labeled data for training, where the diagnosis accuracy may decrease under variable operating conditions. To solve this problem, based on the domain adaptation theory, this paper proposes a novel fault diagnosis framework called subdomain adaptive adversarial network (SAAN) to improve diagnosis accuracy when dealing with variable operating conditions. The proposed method not only makes full use of feature extraction capabilities of dynamic convolution, but also draws on the game idea of generative adversarial networks where feature generator and classifier compete with each other. In addition, the decision boundary of each category is utilized to contribute to accurate classification. Furthermore, the local maximum mean discrepancy (LMMD) is also introduced in the generative network, and the fine-grained information of each category is considered to align the corresponding class spaces of the source and target domains, reducing model classification error near the decision boundary. Finally, the proposed method is verified through two different data sets. Experimental results have shown that the proposed framework has strong generalization performance and desired fault diagnosis accuracy under variable operating conditions.

Key words： fault diagnosis subdomain adaptation dynamic convolution variable operating conditions

收稿日期: 2021-10-08 出版日期: 2022-06-15

引用本文:

周华锋，程培源，邵思羽，赵玉伟. 基于子域自适应对抗网络的轴承故障诊断[J]. 振动与冲击, 2022, 41(11): 114-122.
ZHOU Huafeng, CHENG Peiyuan, SHAO Siyu, ZHAO Yuwei. Bearing fault diagnosis based on subdomain adaptive confrontation network. JOURNAL OF VIBRATION AND SHOCK, 2022, 41(11): 114-122.

链接本文:

http://jvs.sjtu.edu.cn/CN/ 或 http://jvs.sjtu.edu.cn/CN/Y2022/V41/I11/114

[1] Yan J，Meng Y, Lu L, et al. Industrial Big Data in an Industry 4.0 Environment: Challenges, Schemes and Applications for Predictive Maintenance[J]. IEEE Access, 2017:1.
[2] 雷亚国, 贾峰, 孔德同, 等.大数据下机械智能故障诊断的机遇与挑战[J].机械工程学报, 2018,054(005):94-104.
LEI Yaguo,JIA Feng,KONG Detong,et al. Opportunities and Challenges of Machinery Intelligent Fault Diagnosis in Big Data Era[J]. Journal of Mechanical Engineering, 2018,054(005):94-104.
[3] Chen X, Zhang B, Gao D. Bearing fault diagnosis base on multi-scale CNN and LSTM model[J]. Journal of Intelligent Manufacturing, 2021,32.
[4] Guo X, Shen C, Liang C. Deep Fault Recognizer: An Integrated Model to Denoise and Extract Features for Fault Diagnosis in Rotating Machinery[J]. Applied Sciences, 2016,7(1):41.
[5] 赵敬娇, 赵志宏, 杨绍普.基于残差连接和1D-CNN的滚动轴承故障诊断研究[J].振动与冲击,2021,40(10):6.
ZHAO Jingjiao,ZHAO Zhihong,YANG Shaopu.Rolling bearing fault diagnosis based on residual connection and 1D-CNN[J]. Journal of Vibration and Shock, 2021, 40(10): 6.
[6] 张安安, 黄晋英, 冀树伟, 等.基于卷积神经网络图像分类的轴承故障模式识别[J].振动与冲击, 2020,39(04):165-171.
ZHANG An’ an, HUANG Jinying, JI Shuwei, et al. Bearing fault pattern recognition based on image classification with CNN[J]. Journal of Vibration and Shock, 2020, 39(04):165-171.
[7] Wang H, Xu J, Yan R, et al. Intelligent Bearing Fault Diagnosis Using Multi-Head Attention-Based CNN[J]. Procedia Manufacturing, 2020,49:112-118.
[8] Lei Y, Yang B, Jiang X, et al. Applications of machine learning to machine fault diagnosis: A review and roadmap[J]. MECHANICAL SYSTEMS AND SIGNAL PROCESSING, 2020,138.
[9] Yan R, Shen F, Sun C, et al. Knowledge Transfer for Rotary Machine Fault Diagnosis[J]. IEEE SENSORS JOURNAL, 2020,20(15):8374-8393.
[10] Cheng C, Zhou B, Ma G, et al. Wasserstein Distance based Deep Adversarial Transfer Learning for Intelligent Fault Diagnosis with Unlabeled or Insufficient Labeled Data[J]. Neurocomputing, 2020,409(2).
[11] Han T, Liu C, Yang W, et al. Deep Transfer Network with Joint Distribution Adaptation: A New Intelligent Fault Diagnosis Framework for Industry Application[J]. ISA Transactions, 2018.
[12] Qian W, Li S, Wang J. A New Transfer Learning Method and Its Application on Rotating Machine Fault Diagnosis under Variant Working Conditions[J]. IEEE Access, 2018:1.
[13] An Z, Li S, Wang J, et al. Generalization of deep neural network for bearing fault diagnosis under different working conditions using multiple kernel method[J]. Neurocomputing, 2019,352.
[14] Wang X, He H, Li L. A Hierarchical Deep Domain Adaptation Approach for Fault Diagnosis of Power Plant Thermal System[J]. IEEE Transactions on Industrial Informatics, 2019:1.
[15] Li Y, Song Y, Jia L, et al. Intelligent Fault Diagnosis by Fusing Domain Adversarial Training and Maximum Mean Discrepancy via Ensemble Learning[J]. IEEE Transactions on Industrial Informatics, 2020,PP(99):1.
[16] Cao X, Wang Y, Chen B, et al. Domain-adaptive intelligence for fault diagnosis based on deep transfer learning from scientific test rigs to industrial applications[J]. Neural Computing and Applications, 2020,33(2).
[17] ZHU Y, ZHUANG F, WANG J, et al. Deep subdomain adaptation network for image classification[J].IEEE Transactions on Neural Networks and Learning Systems, 2021,32(4):1713-1722.
[18] Chen Y, Dai X, Liu M, et al. Dynamic Convolution: Attention Over Convolution Kernels: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2020[C].
[19] Case Western Reserve University. Case Western Reserve University (CWRU) Bearing Data Center. [EB/OL]. Available:https://csegroups.case.edu/bearingdatacenter/pages/download-data-file/.
[20] Sun B, Saenko K. Deep CORAL: Correlation Alignment for Deep Domain Adaptation[J]. Springer International Publishing, 2016.
[21] LONG M, CAO Y, CAO Z, et al. Transferable representation learning with deep adaptation networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2019,41(12):3071-3085.
[22] LONG M, ZHU H, WANG J, et al. Deep transfer learning with joint adaptation networks[C]// 2016 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR),2016.
[23] Ganin Y, Ajakan H, Germain P, et al. Domain-adversarial training of neural networks[J]. Journal of Machine Learning Research, vol. 17, no. 1, pp. 2030–2096, 2015.
[24] LONG M, CAO Z, WANG J, et al. Conditional adversarial domain adaptation[J]. Advances in Neural Information Processing Systems, 2018 ： 1640–1650.
[25] Shao S, McAleer S, Yan R, et al. Highly Accurate Machine Fault Diagnosis Using Deep Transfer Learning[J]. IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2019,15(4):2446-2455.