Application of Wavelet Packet Transform for Detecting Bearing Issues

Abstract

Vibration condition monitoring analysis is one of the important methods for detecting defects in motors. The vibration data related to the outer race fault of an induction motor (IM) is collected from public domain. The main motive is to observe and characterize outer race defect. Different wavelet transform is mainly used. Absolute energy of the defective condition is compared with the normal state of the motor by the topology of discrete wavelet-transform (DWT). The proper nodes in wavelet packet transform (WPT) can be determined by the best basis selection (BBS) method. The absolute energy of selected nodes is further compared with the normal condition for detecting fault. The sub-band or the node corresponding to the DWT and the WPT analysis is mainly utilized for identifying the proper frequency components in characterizing the defect severity. In the case of DWT-based analysis, the sub-band D<inf>4</inf> (1.5 kHz to 3 kHz) showed a positive energy trend with an increase in defect severity. In WPT, nodes 33 (1.5 kHz to 2.25 kHz) and 34 (2.25 kHz to 3 kHz) showed a positive trend with increased defect severity. Although the change in energy in node 33 is more dominant than in node 34, it can be said that the frequency range corresponding to node 33 (1.5 kHz to 2.25 kHz) can characterize the outer race defect severities. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.