State Prediction of Bearing Based on Relevance Vector Regression Algorithm with RBF Kernel

The scientific and accurate prediction for state of bearing is the key to ensure its safe operation. A rotating bearing monitoring system was presented in U.S. Patent 7606673 and a bearing condition monitoring apparatus was presented in U.S. Patent 8229682, however, the system or apparatus lacks state prediction function of bearing. State prediction of bearing based on relevance vector regression algorithm with RBF kernel is proposed in this paper. Kurtosis of bearing vibration signal can excellently reflect the state of bearing, so the future state of bearing can be excellently reflected by predicting the kurtosis of bearing vibration signal. Thus, kurtosis prediction of bearing vibration signal based on relevance vector regression algorithm with RBF kernel is studied. Finally, the experiments are adopted to demonstrate the feasibility of the proposed method for state prediction of bearing.