The application of wavelet transform and artificial neural networks in machinery fault diagnosis

Abstract

The wavelet transform and artificial neural networks (ANNs) are briefly described. Then both of them are applied comprehensively to machinery fault diagnosis. The wavelet transform is used to pre-process data and extract feature vectors. ANNs are used to identify fault types. Using the wavelet transform, the dimension of the feature vector is greatly decreased and the noises are restrained as well. Thus the construction of the ANNs is simplified and the calculation speed is raised without lowering accuracy. For comparison, two types of features are extracted. Such a diagnosing measure is proved to be efficient by an experiment at the end of the paper.