A Walsh Transform Method to Detect Faults in Mechanical Systems

Abstract

The presence of periodical impulses in vibration signals usually indicates the occurrence of faults in mechanical systems. Unfortunately, in the working condition, detecting the faults in mechanical systems is a difficult job because they are rather weak and often interfered by heavy noise. Therefore, noise suppression is one of the most important tasks in the fault diagnosis for mechanical systems. We proposed a Walsh transform (WT) based method to de-noise the noisy vibration signals. The WT could decompose a noisy signal into a set of rectangular waveforms which are called Walsh coefficients. After filter these coefficients with thresholdings, we reconstruct the signal though inverse Walsh transform (IWT). Finally, the simulation and experimental investigations of the compound faults in roller bearings are implemented to illustrate the feasibility of the WT method.