Analysis of the sound signal to fault detection of bearings based on Variational Mode Decomposition

Abstract

Damage detection in rotating machines is well established for vibration signals. Unfortunately, there are situations, where usage of vibration is not possible. Then, acoustic signal could be used instead. Unfortunately, usually acoustic signal are more noisy and require special treatment for obtain successful damage detection. In the paper we propose to use Variational mode decomposition (VMD) to omit noise for finding de-noise signal. We use vibration data to validate acoustic signal based procedure. The experiment was done on test rig with damaged bearings.