Analysis of rolling bearing fault diagnosis based on

EMD and kurtosis Hilbert envelope demodulation

 

ZHOU Hao, JIA Minping

(School of Mechanical Engineering, Southeast University, Nanjing 211189, China)

 

Abstract: Aiming at solving the problem that fast fourier transformation (FFT) is hardly applied to extract the characteristics frequency of the bearing′s signal, a fault diagnosis method based on empirical mode decomposition (EMD), kurtosis and Hilbert demodulation were proposed. Firstly, EMD was used to decompose the vibration signal into the intrinsic mode function (IMF). Then, some of the IMFs selected by the rule of kurtosis were used to recombine the new vibration signal. At last, the Hilbert envelope demodulation was adopted with the new signal to detect the fault information. Through analyzing the simulation signal and the inner vibration signal of fault rolling bearing respectively, the characteristics frequency could be clearly extracted. The results indicate that the proposed method is effective in extracting the bearings′ fault information and could be used in rolling bearings fault diagnosis.

Key words: fault diagnosis; rolling bearing; empirical mode decomposition(EMD); kurtosis coefficient; Hilbert envelope demodulation