Using PCA with LVQ, RBF, MLP, SOM and Continuous Wavelet Transform for Fault Diagnosis of Gearboxes

A new method based on principal component analysis (PCA) and artificial neural networks (ANN) is proposed for fault diagnosis of gearboxes. Firstly the six different base wavelets are considered, in which three are from real valued and other three from complex valued. Two wavelet selection criteria Maximum Energy to Shannon Entropy ratio and Maximum Relative Wavelet Energy are used and compared to select an appropriate wavelet for feature extraction; next, the continuous wavelet coefficients (CWC) are evaluated for some different scales. As a new method, the optimal range of wavelet scales is selected based on the maximum energy to Shannon entropy ratio criteria and consequently feature vectors are reduced. In addition, energy and Shannon entropy of the wavelet coefficients are used as two new features along with other statistical parameters as input of the classifier. To prevent the curse of dimensionality problem, the principal component analysis applies to this set of features. Finally, the gearbox faults are classified using these statistical features as input to machine learning techniques. Four artificial neural networks are used for faults classifications. The test result showed that the MLP identified the fault categories of gearbox more accurately for both real wavelet and complex wavelet and has a better diagnosis performance as compared to the RBF, LVQ and SOM.