Benchmarking domain adaptation techniques for bearing fault diagnosis

• Victor Afonso Bauler UFSC
• Júlio A. Cordioli UFSC
• Danilo Braga Dynamox SA
• Danilo Silva UFSC

DOI: https://doi.org/10.5753/eniac.2024.244961

Resumo

O avanço tecnológico permitiu o monitoramento simultâneo de milhares de máquinas, possibilitando o treinamento de algoritmos de aprendizado de máquina para realizar a detecção e o diagnóstico de falhas de forma automatizada. Contudo, um dos principais desafios tem sido a falta de dados rotulados, além do fato de que um mesmo tipo de máquina pode ser exposto a diferentes condições de operação, alterando a distribuição dos atributos dos sinais adquiridos. Uma possível solução é utilizar técnicas de adaptação de domínio, que podem melhorar o desempenho dos modelos em novos tipos de máquina, mesmo sem dados rotulados. No entanto, muitos dos artigos apresentados na literatura enfrentam problemas de vazamento de dados, o que impos sibilita a comparação entre técnicas, pois o desempenho dos modelos é irrealista. Diante disso, este artigo propõe uma metodologia para comparar técnicas de adaptação de domínio, evitando esse problema. Foram implementadas oito técnicas de adaptação de domínio e avaliadas em dois conjuntos de dados, resultando em ganhos significativos em um dos cenários de avaliação.

Palavras-chave: Redes Neurais Artificiais, Aprendizado Profundo, Aprendizado de Máquina, Adaptação de domínio

Referências

Bauler, V., Cordioli, J., Braga, D., and Silva, D. (2023). Comparison of traditional vibration analysis techniques and machine learning models for bearing fault detection. In Proceedings of the 27th International Congress of Mechanical Engineering. ABCM.

Ganin, Y., Ustinova, E., Ajakan, H., Germain, P., Larochelle, H., Laviolette, F., March, M., and Lempitsky, V. (2016). Domain-adversarial training of neural networks. Journal of Machine Learning Research, 17(59):1–35.

Hakim, M., Omran, A. A. B., Ahmed, A. N., Al-Waily, M., and Abdellatif, A. (2023). A systematic review of rolling bearing fault diagnoses based on deep learning and transfer learning: Taxonomy, overview, application, open challenges, weaknesses and recommendations. Ain Shams Engineering Journal, 14(4):101945.

He, K., Zhang, X., Ren, S., and Sun, J. (2016). Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition, pages 770–778.

Hendriks, J., Dumond, P., and Knox, D. A. (2022). Towards better benchmarking using the cwru bearing fault dataset. Mechanical Systems and Signal Processing, 169:108732.

Jiao, J., Zhao, M., Lin, J., and Liang, K. (2020). Residual joint adaptation adversarial network for intelligent transfer fault diagnosis. Mechanical Systems and Signal Processing, 145:106962.

Kapoor, S. and Narayanan, A. (2023). Leakage and the reproducibility crisis in machine-learning-based science. Patterns, 4(9):100804.

Kingma, D. P. and Ba, J. (2014). Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980.

Lei, Y., Yang, B., Jiang, X., Jia, F., Li, N., and Nandi, A. K. (2020). Applications of machine learning to machine fault diagnosis: A review and roadmap. Mechanical Systems and Signal Processing, 138:106587.

Li, S., Bu, R., Li, S., Liu, C. H., and Huang, K. (2024). Principal properties attention matching for partial domain adaptation in fault diagnosis. IEEE Transactions on Instrumentation and Measurement, 73:1–12.

Li, Y., Wang, N., Shi, J., Hou, X., and Liu, J. (2018). Adaptive batch normalization for practical domain adaptation. Pattern Recognition, 80:109–117.

Long, M., Cao, Y., Wang, J., and Jordan, M. (2015). Learning transferable features with deep adaptation networks. In Proceedings of the 32nd International Conference on Machine Learning, page 97–105. PMLR.

Long, M., CAO, Z., Wang, J., and Jordan, M. I. (2018). Conditional adversarial domain adaptation. In Advances in Neural Information Processing Systems, volume 31. Curran Associates, Inc.

Long, M., Zhu, H., Wang, J., and Jordan, M. I. (2017). Deep transfer learning with joint adaptation networks. In Proceedings of the 34th International Conference on Machine Learning, page 2208–2217. PMLR.

Neupane, D. and Seok, J. (2020). Bearing fault detection and diagnosis using case western reserve university dataset with deep learning approaches: A review. IEEE Access, 8:93155–93178.

Randall, R. B. (2021). Vibration-Based Condition Monitoring. Wiley, 2 edition.

Rosa, R. K., Braga, D., and Silva, D. (2024). Benchmarking deep learning models for bearing fault diagnosis using the cwru dataset: A multi-label approach. (arXiv:2407.14625). arXiv:2407.14625 [eess].

Smith, W. A. and Randall, R. B. (2015). Rolling element bearing diagnostics using the case western reserve university data: A benchmark study. Mechanical Systems and Signal Processing, 64–65:100–131.

Sun, B. and Saenko, K. (2016). Deep coral: Correlation alignment for deep domain adaptation. In Hua, G. and Jégou, H., editors, Computer Vision – ECCV 2016 Workshops, page 443–450, Cham. Springer International Publishing.

Tzeng, E., Hoffman, J., Saenko, K., and Darrell, T. (2017). Adversarial discriminative domain adaptation. In 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), page 2962–2971.

Wang, Q., Michau, G., and Fink, O. (2019). Domain adaptive transfer learning for fault diagnosis. In 2019 Prognostics and System Health Management Conference (PHM-Paris), page 279–285.

Zhang, S., Su, L., Gu, J., Li, K., Zhou, L., and Pecht, M. (2023). Rotating machinery fault detection and diagnosis based on deep domain adaptation: A survey. Chinese Journal of Aeronautics, 36(1):45–74.

Zhang, S., Zhang, S., Wang, B., and Habetler, T. G. (2020). Deep learning algorithms for bearing fault diagnostics—a comprehensive review. IEEE Access, 8:29857–29881.

Zhang, Y., Liu, T., Long, M., and Jordan, M. (2019). Bridging theory and algorithm for domain adaptation. In Proceedings of the 36th International Conference on Machine Learning, page 7404–7413. PMLR.