A journal of IEEE and CAA , publishes high-quality papers in English on original theoretical/experimental research and development in all areas of automation
Volume 11 Issue 4
Apr.  2024

IEEE/CAA Journal of Automatica Sinica

  • JCR Impact Factor: 11.8, Top 4% (SCI Q1)
    CiteScore: 17.6, Top 3% (Q1)
    Google Scholar h5-index: 77, TOP 5
Turn off MathJax
Article Contents
B. Yang, Y. Lei, X. Li, N. Li, and  A. Nandi,  “Label recovery and trajectory designable network for transfer fault diagnosis of machines with incorrect annotation,” IEEE/CAA J. Autom. Sinica, vol. 11, no. 4, pp. 932–945, Apr. 2024. doi: 10.1109/JAS.2023.124083
Citation: B. Yang, Y. Lei, X. Li, N. Li, and  A. Nandi,  “Label recovery and trajectory designable network for transfer fault diagnosis of machines with incorrect annotation,” IEEE/CAA J. Autom. Sinica, vol. 11, no. 4, pp. 932–945, Apr. 2024. doi: 10.1109/JAS.2023.124083

Label Recovery and Trajectory Designable Network for Transfer Fault Diagnosis of Machines With Incorrect Annotation

doi: 10.1109/JAS.2023.124083
Funds:  This work was supported in part by the National Key R&D Program of China (2022YFB3402100), the National Science Fund for Distinguished Young Scholars of China (52025056), the National Natural Science Foundation of China (52305129), the China Postdoctoral Science Foundation (2023M732789), the China Postdoctoral Innovative Talents Support Program (BX20230290), the Open Foundation of Hunan Provincial Key Laboratory of Health Maintenance for Mechanical Equipment (2022JXKFJJ01), and the Fundamental Research Funds for Central Universities
More Information
  • The success of deep transfer learning in fault diagnosis is attributed to the collection of high-quality labeled data from the source domain. However, in engineering scenarios, achieving such high-quality label annotation is difficult and expensive. The incorrect label annotation produces two negative effects: 1) the complex decision boundary of diagnosis models lowers the generalization performance on the target domain, and 2) the distribution of target domain samples becomes misaligned with the false-labeled samples. To overcome these negative effects, this article proposes a solution called the label recovery and trajectory designable network (LRTDN). LRTDN consists of three parts. First, a residual network with dual classifiers is to learn features from cross-domain samples. Second, an annotation check module is constructed to generate a label anomaly indicator that could modify the abnormal labels of false-labeled samples in the source domain. With the training of relabeled samples, the complexity of diagnosis model is reduced via semi-supervised learning. Third, the adaptation trajectories are designed for sample distributions across domains. This ensures that the target domain samples are only adapted with the pure-labeled samples. The LRTDN is verified by two case studies, in which the diagnosis knowledge of bearings is transferred across different working conditions as well as different yet related machines. The results show that LRTDN offers a high diagnosis accuracy even in the presence of incorrect annotation.

     

  • loading
  • [1]
    Y. Lei, N. Li, and X. Li, Big Data-Driven Intelligent Fault Diagnosis and Prognosis for Mechanical Systems. Berlin, Germany: Springer, Oct. 2022.
    [2]
    W. M. Kouw and M. Loog, “A review of domain adaptation without target labels,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 43, no. 3, pp. 766–785, Mar. 2021. doi: 10.1109/TPAMI.2019.2945942
    [3]
    H. Chen, H. Luo, B. Huang, B. Jiang, and O. Kaynak, “Transfer learning-motivated intelligent fault diagnosis designs: A survey, insights, and perspectives,” IEEE Trans. Neural Networks and Learning Systems, 2023
    [4]
    M. Azamfar, X. Li, and J. Lee, “Intelligent ball screw fault diagnosis using a deep domain adaptation methodology,” Mechanism and Machine Theory, vol. 151, p. 103932, Sept. 2020. doi: 10.1016/j.mechmachtheory.2020.103932
    [5]
    X. Wang, X. Liu, and Y. Li, “An incremental model transfer method for complex process fault diagnosis,” IEEE/CAA J. Autom. Sinica, vol. 6, no. 5, pp. 1268–1280, Sept. 2019. doi: 10.1109/JAS.2019.1911618
    [6]
    S. Schwendemann, Z. Amjad, and A. Sikora, “Bearing fault diagnosis with intermediate domain based layered maximum mean discrepancy: A new transfer learning approach,” Engineering Applications of Artificial Intelligence, vol. 105, p. 104415, Oct. 2021. doi: 10.1016/j.engappai.2021.104415
    [7]
    B. Rezaeianjouybari and Y. Shang, “A novel deep multi-source domain adaptation framework for bearing fault diagnosis based on feature-level and task-specific distribution alignment,” Measurement, vol. 178, p. 109359, Jun. 2021. doi: 10.1016/j.measurement.2021.109359
    [8]
    B. Sun and K. Saenko, “Deep coral: Correlation alignment for deep domain adaptation,” in Proc. Computer Vision–European Conf. Computer Vision Workshops, Amsterdam, Netherlands, 2016, pp. 443–450.
    [9]
    Q. Wang, G. Michau, and O. Fink, “Domain adaptive transfer learning for fault diagnosis,” in Proc. Prognostics and System Health Management Conf., Paris, France, 2019, pp. 279–285.
    [10]
    J. Lee, M. Kim, J. Ko, J. Jung, K. Sun, and B. Youn, “Asymmetric inter-intra domain alignments (AIIDA) method for intelligent fault diagnosis of rotating machinery,” Reliability Engineering &System Safety, vol. 218, p. 108186, Feb. 2022.
    [11]
    B. Wang, P. Baraldi, and E. Zio, “Deep multi-adversarial conditional domain adaptation networks for fault eiagnostics of industrial equipment,” IEEE Trans. Ind. Informatics, vol. 19, no. 8, pp. 8841–8851, Aug. 2023. doi: 10.1109/TII.2022.3222400
    [12]
    I. Goodfellow, J. Pouget-Abadie, M. Mirza, et al., “Generative adversarial networks,” Communi. ACM, vol. 63, no. 11, pp. 139–144, Oct. 2020. doi: 10.1145/3422622
    [13]
    M. Ghorvei, M. Kavianpour, M. T. Beheshti, and A. Ramezani, “An unsupervised bearing fault diagnosis based on deep subdomain adaptation under noise and variable load condition,” Measurement Science and Tech., vol. 33, no. 2, p. 25901, Dec. 2021.
    [14]
    M. Ghorvei, M. Kavianpour, M. T. Beheshti, and A. Ramezani, “Spatial graph convolutional neural network via structured subdomain adaptation and domain adversarial learning for bearing fault diagnosis,” Neurocomputing, vol. 517, pp. 44–61, Jan. 2023. doi: 10.1016/j.neucom.2022.10.057
    [15]
    M. Kavianpour, A. Ramezani, and M. T. Beheshti, “A class alignment method based on graph convolution neural network for bearing fault diagnosis in presence of missing data and changing working conditions,” Measurement, vol. 199, p. 111536, Aug. 2022. doi: 10.1016/j.measurement.2022.111536
    [16]
    B. Yang, Y. Lei, X. Li, and C. Roberts, “Deep targeted transfer learning along designable adaptation trajectory for fault diagnosis across different machines,” IEEE Trans. Ind. Electronics, vol. 70, no. 9, pp. 9463–9473, Sept. 2023. doi: 10.1109/TIE.2022.3212415
    [17]
    X. Li, S. Yu, Y. Lei, N. Li, and B. Yang, “Intelligent machinery fault diagnosis with event-based camera,” IEEE Trans. Ind. Informatics, vol. 20, no. 1, pp. 380–389, Jan. 2024. doi: 10.1109/TII.2023.3262854
    [18]
    B. Frenay and M. Verleysen, “Classification in the presence of label noise: A survey,” IEEE Trans. Neural Networks and Learning Systems, vol. 25, no. 5, pp. 845–869, May 2014. doi: 10.1109/TNNLS.2013.2292894
    [19]
    E. Tzeng, J. Hoffman, N. Zhang, K. Saenko, and T. Darrell, “Deep domain confusion: Maximizing for domain invariance,” arXiv preprint arXiv: 1412.3474, 2014.
    [20]
    Y. Lei, B. Yang, X. Jiang, F. Jia, N. Li, and A. K. Nandi, “Applications of machine learning to machine fault diagnosis: A review and roadmap,” Mechanical Systems and Signal Processing, vol. 138, p. 106587, Apr. 2020. doi: 10.1016/j.ymssp.2019.106587
    [21]
    Y. Ganin, E. Ustinova, H. Ajakan, et al., “Domain-adversarial training of neural networks,” J. Machine Learning Research, vol. 17, no. 1, pp. 2096–2030, Jan. 2016.
    [22]
    Z. Liu, L. Jiang, H. Wei, L. Chen, and X. Li, “Optimal transport-based deep domain adaptation approach for fault diagnosis of rotating machine,” IEEE Trans. Instrumentation and Measurement, vol. 70, p. 3508912, Jan. 2021.
    [23]
    B. B. Damodaran, B. Kellenberger, R. Flamary, D. Tuia, and N. Courty, “DeepJDOT: Deep joint distribution optimal transport for unsupervised domain adaptation,” in Proc. European Conf. Computer Vision, Munich, Germany, 2018, pp. 467–483.
    [24]
    H. Song, M. Kim, D. Park, Y. Shin, and J. G. Lee, “Learning from noisy labels with deep neural networks: A survey,” IEEE Trans. Neural Networks and Learning Systems, vol. 34, no. 11, pp. 8135–8153, Nov. 2023. doi: 10.1109/tnnls.2022.3152527
    [25]
    J. Goldberger and E. Ben-Reuven, “Training deep neural-networks using a noise adaptation layer,” in Proc. 5th Int. Conf. Learning Representations, Toulon, France, 2017.
    [26]
    T. Liu and D. Tao, “Classification with noisy labels by importance reweighting,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 38, no. 3, pp. 447–461, Mar. 2016. doi: 10.1109/TPAMI.2015.2456899
    [27]
    L. Jiang, Z. Zhou, T. Leung, L. Li, and F. Li, “MentorNet: Learning data-driven curriculum for very deep neural networks on corrupted labels,” in Proc. 35th Int. Conf. Machine Learning, Stockholmsmässan, Stockholm Sweden, 2018, pp. 2304–2313.
    [28]
    B. Han, Q. Yao, X. Yu, et al., “Co-teaching: Robust training of deep neural networks with extremely noisy labels,” in Proc. 32nd Int. Conf. Neural Information Processing Systems, Montréal, Canada, 2018, pp. 8536–8546.
    [29]
    Y. Zhang, Y. Wei, Q. Wu, et al., “Collaborative unsupervised domain adaptation for medical image diagnosis,” IEEE Trans. Image Processing, vol. 29, pp. 7834–7844, Jul. 2020. doi: 10.1109/TIP.2020.3006377
    [30]
    Y. Shu, Z. Cao, M. Long, and J. Wang, “Transferable curriculum for weakly-supervised domain adaptation,” in Proc. AAAI Conf. Artificial Intelligence, Hawaii, USA, 2019, pp. 4951–4958.
    [31]
    Q. Yu, A. Hashimoto, and Y. Ushiku, “Divergence optimization for noisy universal domain adaptation,” in Proc. IEEE/CVF Conf. Computer Vision and Pattern Recognition, Electr Network, 2021, pp. 2515–2524.
    [32]
    G. Peyré and M. Cuturi, “Computational optimal transport: With applications to data science,” Foundations and Trends® in Machine Learning, vol. 11, no. 5–6, pp. 355–607, Feb. 2019.
    [33]
    M. K. Afzal, J. M. Adam, H. M. R. Afzal, et al., “Discriminative feature abstraction by deep L-2 hypersphere embedding for 3D mesh CNNs,” Infor. Sciences, vol. 607, pp. 1158–1173, Aug. 2022. doi: 10.1016/j.ins.2022.05.104
    [34]
    J. Han, P. Luo, and X. Wang, “Deep self-learning from noisy labels,” in Proc. IEEE/CVF Conf. Computer Vision and Pattern Recognition, Seoul, South Korea, 2019, pp. 5137–5146.
    [35]
    K. Saito, K. Watanabe, Y. Ushiku, and T. Harada, “Maximum classifier discrepancy for unsupervised domain adaptation,” in Proc. IEEE/CVF Conf. Computer Vision and Pattern Recognition, Salt Lake City, USA, 2018, pp. 3723–3732.
    [36]
    M. Cuturi, “Sinkhorn distances: Lightspeed computation of optimal transport,” in Proc. 26th Int. Conf. Neural Information Processing Systems, Lake Tahoe, USA, 2013, pp. 2292–2300.
    [37]
    J. Zhang, Z. Ding, W. Li, and P. Ogunbona, “Importance weighted adversarial nets for partial domain adaptation,” in Proc. IEEE/CVF Conf. Computer Vision and Pattern Recognition, Salt Lake City, USA, 2018, pp. 8156–8164.
    [38]
    B. Yang, Y. Lei, S. Xu, and C.-G. Lee, “An optimal transport-embedded similarity measure for diagnostic knowledge transferability analytics across machines,” IEEE Trans. Ind. Electronics, vol. 69, no. 7, pp. 7372–7382, Jul. 2022. doi: 10.1109/TIE.2021.3095804

Catalog

    通讯作者: 陈斌, bchen63@163.com
    • 1. 

      沈阳化工大学材料科学与工程学院 沈阳 110142

    1. 本站搜索
    2. 百度学术搜索
    3. 万方数据库搜索
    4. CNKI搜索

    Figures(13)  / Tables(6)

    Article Metrics

    Article views (196) PDF downloads(52) Cited by()

    Highlights

    • A semi-supervised transfer learning framework is presented for intelligent diagnosis
    • Domain adaptation is achieved in the presence of a false-labeled source domain
    • A new indicator is proposed to evaluate and further modify label anomaly
    • The adaptation trajectory to false-labeled samples is avoided by OT theory

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return