Normalized wavelet packets quantifiers for condition monitoring

Normalized wavelet packets quantifiers are proposed and studied as a new tool for condition monitoring. The new quantifiers construct a complete quantitative time-frequency analysis: the Wavelet packets relative energy measures the normalized energy of the wavelet packets node; the Total wavelet packets entropy measures how the normalized energies of the wavelet packets nodes are distributed in the frequency domain; the Wavelet packets node entropy describes the uncertainty of the normalized coefficients of the wavelet packets node. Unlike the feature extraction methods directly using the amplitude of wavelet coefficients, the new quantifiers are derived from probability distributions and are more robust in diagnostic applications. By applying these quantifiers to Acoustic Emission signals from faulty bearings of rotating machines, our study shows that both localized defects and advanced contamination faults can be successfully detected and diagnosed if the appropriate quantifier is chosen. The Bayesian classifier is used to quantitatively analyse and evaluate the performance of the proposed quantifiers. We also show that reducing the Daubechies wavelet order or the length of the segment will deteriorate the performance of the quantifiers. A two-dimensional diagnostic scheme can also help to improve the diagnostic performance but the improvements are only significant when using lower wavelet orders. Crown Copyright (C) 2008 Published by Elsevier Ltd. All rights reserved.