45 Corresponding author email : a.damou@univ-boumerdes.dz Machine Learning-Enhanced Gear Fault Detection and Classification via Wavelet Packet Transform Analysis

Abstract

Gear fault detection, diagnosis, and classification are critical yet challenging tasks due to the similarity in spectral features among various gear fault types. This study presents an automated approach for detecting, identifying, and classifying gear defects by leveraging Wavelet Packet Transform (WPT) for signal decomposition and feature extraction.