Representation Learning–Based Fault Diagnosis of Angular Contact Ball Bearings for Machine Tool Spindles
##plugins.themes.bootstrap3.article.main##
##plugins.themes.bootstrap3.article.sidebar##
Abstract
As a precision mechanical component designed to reduce friction between moving parts, the rolling bearing is widely employed across various industries. Predicting the remaining useful life (RUL) of bearings is critical for preventing unexpected failures and ensuring safe and reliable equipment operation. Equally important is the role of lubrication, which not only supports proper bearing performance but also ensures the accurate operation of CNC spindles. The primary objective of predicting lubricant degradation is therefore to estimate the time at which the lubricant no longer fulfills its intended function
In this study, bearing defect frequencies and lubricant anomalies are investigated through vibration-signal analysis under different operating conditions using a representation learning approach. To be more specific, a representation framework is employed to reconstruct input signals. A target frequency band, extending from three times the rotational frequency to 10,000 Hz, is first defined. Spectral root mean square (Spectral RMS) features are then extracted from the low-to-middle frequency portion of this band, where defect-induced impulsive excitations are effectively captured. The framework is trained using labeled healthy-bearing datasets to establish an anomaly-detection threshold, which is subsequently applied during the testing phase to evaluate reconstruction loss. The threshold is defined as the 95th percentile of the reconstruction means square error obtained during training. This criterion enables the identification of anomalies in labeled fault-condition data. The effectiveness of the proposed method is demonstrated by employing two practical datasets. The results indicate that the proposed method effectively detects anomalies in unseen data and achieves robust diagnostic performance
How to Cite
##plugins.themes.bootstrap3.article.details##
Bearing fault detection, Lubricant failure detection, Multi-head Attention, Representation learning, Spectral Root-mean-square
Gong, F., Ma, P., Zhang, H., Wang, C., Li, X., & Wu, Y. (2025). Rolling bearings remaining useful life estimation using digital twin and physics-informed methods with uncertainty quantification. Engineering Applications of Artificial Intelligence, 154, 111070. doi: 10.1016/j.engappai.2025.111070
Huang, C. G., Huang, H. Z., Li, Y. F., & Peng, W. (2021). A novel deep convolutional neural network-bootstrap integrated method for RUL prediction of rolling bearing. Journal of Manufacturing Systems, 61, 757–772. doi: 10.1016/j.jmsy.2021.03.012
Kim, S., Seo, Y. H., & Park, J. (2024). Transformer-based novel framework for remaining useful life prediction of lubricant in operational rolling bearings. Reliability Engineering & System Safety, 251, 110377. doi: 10.1016/j.ress.2024.110377
Lee, J., Qiu, H., Yu, G., Lin, J., & Rexnord Technical Services. (2007). Bearing data set. NASA Ames Prognostics Data Repository, IMS, University of Cincinnati.
Li, H., Zhang, Z., Li, T., & Si, X. (2024). A review on physics-informed data-driven remaining useful life prediction: Challenges and opportunities. Mechanical Systems and Signal Processing, 209, 111120. doi: 10.1016/j.ymssp.2024.111120
Mei, J., Jia, J., Zeng, R., Zhou, B., & Zhao, H. (2016). A multi-order FRFT self-adaptive filter based on segmental frequency fitting and early fault diagnosis in gears. Measurement: Journal of the International Measurement Confederation, 91, 532–540. doi: 10.1016/j.measurement.2016.05.092
Shang, Y., Tang, X., Zhao, G., Jiang, P., & Ran Lin, T. (2022). Remaining life prediction of rolling element bearings based on a bidirectional gated recurrent unit and convolutional neural network. Measurement: Journal of the International Measurement Confederation, 202, 111893. doi: 10.1016/j.measurement.2022.111893
Shen, S., Lu, H., Sadoughi, M., Hu, C., Nemani, V., Thelen, A., ... Kenny, S. (2021). A physics-informed deep learning approach for bearing fault detection. Engineering Applications of Artificial Intelligence, 103, 104295. doi: 10.1016/j.engappai.2021.104295
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., ... Polosukhin, I. (2017). Attention is all you need. Advances in Neural Information Processing Systems, 30. Retrieved from https://arxiv.org/pdf/1706.03762
This work is licensed under a Creative Commons Attribution 3.0 Unported License.
The Prognostic and Health Management Society advocates open-access to scientific data and uses a Creative Commons license for publishing and distributing any papers. A Creative Commons license does not relinquish the author’s copyright; rather it allows them to share some of their rights with any member of the public under certain conditions whilst enjoying full legal protection. By submitting an article to the International Conference of the Prognostics and Health Management Society, the authors agree to be bound by the associated terms and conditions including the following:
As the author, you retain the copyright to your Work. By submitting your Work, you are granting anybody the right to copy, distribute and transmit your Work and to adapt your Work with proper attribution under the terms of the Creative Commons Attribution 3.0 United States license. You assign rights to the Prognostics and Health Management Society to publish and disseminate your Work through electronic and print media if it is accepted for publication. A license note citing the Creative Commons Attribution 3.0 United States License as shown below needs to be placed in the footnote on the first page of the article.
First Author et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 United States License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.