Machine Fault Diagnosis Using Mutual Information and Informative Wavelet

Abstract

This paper deals with an application of wavelets for feature extraction and classification of machine faults. The statistical approach referred to as informative wavelet algorithm is utilized to generate wavelets and subsequent coefficients that are used as feature variables for the classification and diagnosis of machine faults. Informative wavelets are referred to classes of functions generated from elements of a dictionary of orthogonal bases, such as wavelet packet dictionary. Training data are used to construct probability distributions required for the computation of the entropy and mutual information. In our data analysis, we have used machine data acquired from a single cylinder engine under a series of induced faults in a test environment. The objective of the experiment was to evaluate the performance of the informative wavelet algorithm in classifying faults using real-world machine data and to examine the extent to which the results were influenced by different analyzing wavelets chosen for data analysis. The correlation structure of the informative wavelets as well as coefficient matrix are also examined.