Virtual Sensor for Real-Time Bearing Load Prediction Using Heterogeneous Temporal Graph Neural Networks

##plugins.themes.bootstrap3.article.main##

##plugins.themes.bootstrap3.article.sidebar##

Published Jun 27, 2024
Mengjie Zhao
Cees Taal Stephan Baggerohr Olga Fink

Abstract

Accurate bearing load monitoring is essential for their Prognostics and Health Management (PHM), enabling damage assessment, wear prediction, and proactive maintenance. While bearing sensors are typically placed on the bearing housing, direct load monitoring requires sensors inside the bearing itself.
Recently introduced sensor rollers enable direct bearing load monitoring but are constrained by their battery life.
Data-driven virtual sensors can learn from sensor roller data collected during a battery's lifetime to map operating conditions to bearing loads.
Although spatially distributed bearing sensors offer insights into load distribution (e.g., correlating temperature with load), traditional machine learning algorithms struggle to fully exploit these spatial-temporal dependencies. To address this gap, we introduce a graph-based virtual sensor that leverages Graph Neural Networks (GNNs) to analyze spatial-temporal dependencies among sensor signals, mapping existing measurements (temperature, vibration) to bearing loads. Since temperature and vibration signals exhibit vastly different dynamics, we propose Heterogeneous Temporal Graph Neural Networks (HTGNN), which explicitly models these signal types and their interactions for effective load prediction.
Our results demonstrate that HTGNN outperforms Convolutional Neural Networks (CNNs), which struggle to capture both spatial and heterogeneous signal characteristics. These findings highlight the importance of capturing the complex spatial interactions between temperature, vibration, and load.

How to Cite

Zhao, M., Taal, C., Baggerohr, S., & Fink, O. (2024). Virtual Sensor for Real-Time Bearing Load Prediction Using Heterogeneous Temporal Graph Neural Networks. PHM Society European Conference, 8(1), 8. https://doi.org/10.36001/phme.2024.v8i1.3998
Abstract 442 | PDF Downloads 207

##plugins.themes.bootstrap3.article.details##

Keywords

Virtual Sensor, Bearing, Load Prediction, Virtual Sensing, Soft Sensing, Graph Neural Networks, Heterogeneous Sensor Networks, Vibration Analysis, WInd Turbine Bearings, Heterogeneous Graph Neural Network, Time Series Analysis, Multivariate Time Series

References
Baggerohr, S. (2023). On the detection of rolling contact fatigue in large bearings using roller embedded sensors. In Surveillance, vibrations, shock and noise. Bilbao, J., Lourens, E.-M., Schulze, A., & Ziegler, L. (2022). Virtual sensing in an onshore wind turbine tower using a gaussian process latent force model. Data-Centric Engineering, 3, e35. Brody, S., Alon, U., & Yahav, E. (2022). How attentive are graph attention networks? In International conference on learning representations (iclr). Chao, M. A., Kulkarni, C., Goebel, K., & Fink, O. (2022). Fusing physics-based and deep learning models for prognostics. Reliability Engineering & System Safety, 217, 107961. Deng, A., & Hooi, B. (2021). Graph neural network-based anomaly detection in multivariate time series. In Proceedings of the aaai conference on artificial intelligence (Vol. 35, pp. 4027–4035). Dimitrov, N., & G¨oc¸men, T. (2022). Virtual sensors for wind turbines with machine learning-based time series models. Wind Energy, 25(9), 1626–1645. Gilmer, J., Schoenholz, S. S., Riley, P. F., Vinyals, O., & Dahl, G. E. (2017). Neural message passing for quantum chemistry. In International conference on machine learning (pp. 1263–1272). Harris, T. A., & Kotzalas, M. N. (2006). Essential concepts of bearing technology. CRC press. Hou, Y., & Wang, X. (2021). Measurement of load distribution in a cylindrical roller bearing with an instrumented housing: Finite element validation and experimental study. Tribology International, 155, 106785. Jin, M., Koh, H. Y., Wen, Q., Zambon, D., Alippi, C., Webb, G. I., ... Pan, S. (2023). A survey on graph neural networks for time series: Forecasting, classification, imputation, and anomaly detection. arXiv preprint arXiv:2307.03759. Kerst, S., Shyrokau, B., & Holweg, E. (2019). A model-based approach for the estimation of bearing forces and moments using outer ring deformation. IEEE Transactions on Industrial Electronics, 67(1), 461–470. Kipf, T. N., & Welling, M. (2017). Semi-supervised classification with graph convolutional networks. In 5th

international conference on learning representations. Konopka, D., Steppeler, T., Ottermann, R., Pape, F., Dencker, F., Poll, G., & Wurz, M. (2023). Advancements in monitoring of tribological stress in bearings using thinfilm strain gauges. Martin, D., K¨uhl, N., & Satzger, G. (2021). Virtual sensors. Business & Information Systems Engineering, 63, 315–

323.

Morales, G. E., Engelen, P., & Van Nijen, G. (2019). Propagation of large spalls in rolling bearings. Tribology Online, 14(5), 254-266. doi: 10.2474/trol.14.254 Niresi, K. F., Zhao, M., Bissig, H., Baumann, H., & Fink,

O. (2023). Spatial-temporal graph attention fuser for calibration in iot air pollution monitoring systems. In 2023 ieee sensors (pp. 01–04). Peng, D., Wang, H., Liu, Z., Zhang, W., Zuo, M. J., & Chen,

J. (2020). Multibranch and multiscale cnn for fault diagnosis of wheelset bearings under strong noise and variable load condition. IEEE Transactions on Industrial Informatics, 16(7), 4949–4960. Shi, C. (2022). Heterogeneous graph neural networks. Graph Neural Networks: Foundations, Frontiers, and Applications, 351–369. Song, Z., Hackl, C. M., Anand, A., Thommessen, A., Petzschmann, J., Kamel, O., . . . Hauptmann, S. (2023). Digital twins for the future power system: An overview and a future perspective. Sustainability, 15(6), 5259. Wang, G., Jia, Q.-S., Zhou, M., Bi, J., & Qiao, J. (2021). Soft-sensing of wastewater treatment process via deep belief network with event-triggered learning. Neurocomputing, 436, 103–113. Widner, R., & Littmann, W. (1976). Bearing damage analysis. National Bureau of Standard special publication(423), 1. Yang, X., Zheng, Y., Zhang, Y., Wong, D. S.-H., & Yang,

W. (2022). Bearing remaining useful life prediction based on regression shapalet and graph neural network. IEEE Transactions on Instrumentation and Measurement, 71, 1–12. Zhao, M., & Fink, O. (2023). Dyedgegat: Dynamic edge via graph attention for early fault detection in iiot systems. arXiv preprint arXiv:2307.03761.
Section
Technical Papers