Research on fault diagnosis method based on ITD & MED

Abstract

In order to extract the fault characteristics of rolling bearing from complex operating conditions, a fault diagnosis method is proposed based on Intrinsic Time-scale Decomposition (ITD) and Minimum Entropy Deconvolution (MED). Firstly, by applying ITD to decompose vibration signals, a great deal of Proper Rotation (PR) shall be obtained. And those PR containing the most fault information shall be used for signal restructure based on the kurtosis criterion. Then with the use of MED, the restructured signals are able to be reduced and the impact features of those signals shall be enhanced. Finally, the Teager energy operator has been used to calculate the deduction of noise reduction signal and to draw the Teager energy spectrum which can identify the fault features of roll bearing. With the adaptation of this method for fault diagnosis of the rolling bearing, the experimental results have verified the effectiveness of the method. Key Words: Rolling bearing; Minimum entropy deconvolution; ITD; Teager energy operator; Fault diagnosis