Bearing remaining life prediction using Gaussian process regression with composite kernel functions

There is an urgent demand for life prediction of bearing in industry. Effective bearing degradation assessment technique is beneficial to condition based maintenance (CBM). In this paper, Gaussian Process Regression (GPR) is used for remaining bearing life prediction. Three main steps of prediction schedule are presented in details. RMS, Kurtosis and Crest factor are used for feature fusion by self-organizing map (SOM). Minimum Quantization Error (MQE) value derived from SOM is applied to represent the condition of bearing. GPR models with both single and composite covariance functions are presented. After training, new MQE value can be predicted by the GPR model according to previous data points. Experimental results show that composite kernels improve the accuracy and reduce the variance of prediction results. Compared with particle filter (PF), GPR model can predict the remaining life of bearings more accurately.

[1]  Aouni A. Lakis,et al.  Diagnosis of rotor bearings using logical analysis of data , 2011 .

[2]  Bo-Suk Yang,et al.  Application of relevance vector machine and logistic regression for machine degradation assessment , 2010 .

[3]  Teuvo Kohonen,et al.  Self-Organizing Maps , 2010 .

[4]  Lei Guo,et al.  Robust bearing performance degradation assessment method based on improved wavelet packet–support vector data description , 2009 .

[5]  Joseph Mathew,et al.  Rotating machinery prognostics. State of the art, challenges and opportunities , 2009 .

[6]  Takashi Hiyama,et al.  Predicting remaining useful life of rotating machinery based artificial neural network , 2010, Comput. Math. Appl..

[7]  Jay Lee,et al.  Robust performance degradation assessment methods for enhanced rolling element bearing prognostics , 2003, Adv. Eng. Informatics.

[8]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[9]  Enrico Zio,et al.  Condition assessment for the performance degradation of bearing based on a combinatorial feature extraction method , 2014, Digit. Signal Process..

[10]  Simon J. Godsill,et al.  On sequential Monte Carlo sampling methods for Bayesian filtering , 2000, Stat. Comput..

[11]  Nagi Gebraeel,et al.  Residual life predictions from vibration-based degradation signals: a neural network approach , 2004, IEEE Transactions on Industrial Electronics.

[12]  Sirkka-Liisa Jämsä-Jounela,et al.  A process monitoring system based on the Kohonen self-organizing maps , 2003 .

[13]  Lifeng Xi,et al.  Residual life predictions for ball bearings based on self-organizing map and back propagation neural network methods , 2007 .

[14]  Zhengjia He,et al.  Research on bearing life prediction based on support vector machine and its application , 2011 .

[15]  Sankalita Saha,et al.  A Distributed Prognostic Health Management Architecture , 2009 .

[16]  Sankalita Saha,et al.  Distributed prognostic health management with gaussian process regression , 2010, 2010 IEEE Aerospace Conference.

[17]  Peng Wang,et al.  Fault prognostics using dynamic wavelet neural networks , 2001, Artificial Intelligence for Engineering Design, Analysis and Manufacturing.

[18]  Sheng Hong,et al.  Performance Degradation Assessment for Bearing Based on Ensemble Empirical Mode Decomposition and Gaussian Mixture Model , 2014 .

[19]  Xu Chong,et al.  INTELLIGENT ANALYSIS MODEL OF SLOPE NONLINEAR DISPLACEMENT TIME SERIES BASED ON GENETIC-GAUSSIAN PROCESS REGRESSION ALGORITHM OF COMBINED KERNEL FUNCTION , 2009 .