Fault diagnosis of rolling bearing based on empirical mode decomposition and higher order statistics

• Published in: Proceedings of the Institution of Mechanical Engineers. Part C, Journal of mechanical engineering science Journal of mechanical engineering science, 2015-06, Vol.229 (9), p.1630-1638
• Main Author: Cai, Jian-hua
• Format: Article
• Language: English
• Subjects: Bearings; Decomposition; Empirical analysis; Fault diagnosis; Mathematical models; Mechanical engineering; Normal distribution; Race; Roller bearings; Rolling bearings; Simulation; Statistics; Vibration
• ISSN: 0954-4062; 2041-2983
• DOI: 10.1177/0954406214545820

In order to solve the problem of the faulted rolling bearing signal getting easily affected by Gaussian noise, a new fault diagnosis method was proposed based on empirical mode decomposition and high-order statistics. Firstly, the vibration signal was decomposed by empirical mode decomposition and the correlation coefficient of each intrinsic mode function was calculated. These intrinsic mode function components, which have a big correlation coefficient, were selected to estimate its higher order spectrum. Then based on the higher order statistics theory, this method uses higher order spectrum of each intrinsic mode function to reconstruct its power spectrum. And these power spectrums were summed to obtain the primary power spectrum of bearing signal. Finally, fault feature information was extracted from the reconstructed power spectrum. A model, using higher order spectrum to reconstruct power spectrum, was established. Meanwhile, analysis was conducted by using the simulated data and the recorded vibration signals which include inner race, out race, and bearing ball fault signal. Results show that the presented method is superior to traditional power spectrum method in suppressing Gaussian noise and its resolution is higher. New method can extract more useful information compared to the traditional method.