Fault Prediction of Rolling Bearing based on Frequency Domain Analysis 

Jie YUAN

Chongqing Maritime Bureau, Chongqing, 400000, China

Abstract: A failure prognosis model which fuses the frequency domain analysis of vibration signals, the re-liability analysis and BP neural network is proposed. Primarily, the frequency domain analysis of antifriction bearing vibration data is conducted by discrete Fourier transform. And then, the reliability of antifriction bearings is assessed by making use of Weibull distribution based on the maintenance data. Finally, BP neural network is trained and examined by Matlab using sample data.

Keywords: Failure prognosis; Frequency domain analysis; Reliability analysis; BP neural network; Antifric-tion bearing