A Feature Extraction Method Based on Wavelet Packet and Discrete Cosine Transform for Gearbox Fault Diagnosis

摘要

A new feature extraction method based on wavelet packet and discrete cosine transform for gearbox fault diagnosis was proposed in this paper. Firstly, the gearbox vibration signals were de-noise and decomposed into various frequencies width subspace based on wavelet packet. Then a new fault feature parameters based on discrete cosine transform coefficients for each sub-space were proposed to evaluate gearbox fault advancement quantitatively. Finally,we test our method with three faults signals. We found that these parameters can correctly indicate early gearbox fault. For a comparison, the advantages and disadvantages of tradition feature parameter is also discussed and compare two method results. The results show that the proposed method has better diagnosis effect for early fault.