Unscented Kalman Filter with Gaussian Process Degradation Model for Bearing Fault Prognosis



Published Jul 3, 2012
Christoph Anger Robert Schrader Uwe Klingauf


The degradation of rolling-element bearings is mainly stochastic due to unforeseeable influences like short term overstraining, which hampers the prediction of the remaining useful lifetime. This stochastic behaviour is hardly describable with parametric degradation models, as it has been done in the past. Therefore, the two prognostic concepts presented and examined in this paper introduce a nonparametric approach by the application of a dynamic Gaussian Process (GP). The GP offers the opportunity to reproduce a damage course according to a set of training data and thereby also estimates the uncertainties of this approach by means of the GP’s covariance. The training data is generated by a stochastic degradation model that simulates the aforementioned highly stochastic degradation of a bearing fault. For prediction and state estimation of the feature, the trained dynamic GP is combined with the Unscented Kalman Filter (UKF) and evaluated
in the context of a case study. Since this prognostic approach has shown drawbacks during the evaluation, a multiple model approach based on GP-UKF is introduced and evaluated. It is shown that this combination offers an increased prognostic performance for bearing fault prediction.

How to Cite

Anger, C., Schrader, R., & Klingauf, U. (2012). Unscented Kalman Filter with Gaussian Process Degradation Model for Bearing Fault Prognosis. PHM Society European Conference, 1(1). https://doi.org/10.36001/phme.2012.v1i1.1416
Abstract 240 | PDF Downloads 186



remaining useful life (RUL), Data-driven and model-based prognostics, Gaussian Process Model, Bearing Faults, Multiple Model

Antoni, J. (2007). Cyclic spectral analysis of rolling-element bearing signals: Facts and fictions. Journal of Sound and vibration, 304(3-5), 497–529.
Choi, Y., & Liu, C. R. (2006a). Rolling contact fatigue life of finish hard machined surfaces - Part 1. Model development. Wear, 261(5-6), 485–491.
Choi, Y., & Liu, C. R. (2006b). Rolling contact fatigue life of finish hard machined surfaces - Part 2. Experimental verification. Wear, 261(5-6), 492–499.
Daigle, M., & Goebel, K. (2010). Model-Based Prognostics under Limited Sensing.
Julier, S. (2002). The scaled unscented transformation. In American Control Conference, 2002. Proceedings of the 2002 (Vol. 6, pp. 4555–4559).
Ko, J., & Fox, D. (2011). Learning GP-BayesFilters via Gaussian process latent variable models. Autonomous Robots, 30(1), 3–23.
Ko, J., Klein, D., Fox, D., & Haehnel, D. (2007). GPUKF: Unscented Kalman filters with Gaussian process prediction and observation models. In Intelligent Robots and Systems, 2007. IROS 2007. IEEE/RSJ International Conference on (pp. 1901–1907).
Li, X., & Jilkov, V. (2003). A survey of maneuvering target tracking—Part V: Multiple-model methods. In Proc. SPIE Conf. on Signal and Data Processing of Small Targets (pp. 559–581).
Orchard, M. E., & Vachtsevanos, G. J. (2009). A particlefiltering approach for on-line fault diagnosis and failure prognosis. Transactions of the Institute of Measurement and Control, 31(3-4), 221–246.
Orsagh, R., Sheldon, J., & Klenke, C. (2003). Prognostics/diagnostics for gas turbine engine bearings. In Proceedings of IEEE Aerospace Conference.
Rasmussen, C. E., & Williams, C. K. I. (2006). Gaussian Processes for Machine Learning.
Saxena, A., Celaya, J., Balaban, E., Goebel, K., Saha, B., Saha, S., et al. (2008). Metrics for evaluating performance of prognostic techniques. In Prognostics and Health Management, 2008. PHM 2008. International Conference on (pp. 1–17).
Schaab, J. (2011). Trusted health assessment of dynamic systems based on hybrid joint estimation (Als Ms. gedr. ed.). D¨usseldorf: VDI-Verl.
Sturm, A. (1986). W¨alzlagerdiagnose an Maschinen und Anlagen. Ko¨ln: TU¨ V Rheinland. Yu, W. K., & Harris, T. A. (2001). A New Stress-Based Fatigue Life Model for Ball Bearings. Tribology Transactions, 44(1), 11–18.
Technical Papers