Research On Rub And Impact Fault Identification Method Based On Wavelet Packetand Support Vector Machine

The rub fault often occurs between the stators in modern rotating machinery during operation. If the faults are not found in time, it will lead to serious consequences. Therefore an intelligent diagnostic method for rub-impact fault based on wavelet packet and support vector machine is proposed in this paper.Fault Diagnosis Technology is a pattern recognition event to the running machine in essence, and the key lies in the signal analysis and fault features extraction, which has a direct bearing on the accuracy of fault diagnosis and fault the reliability of early warning. Wavelet and wavelet packet analysis method have good time-frequency localization, and can divide the signal at any frequency band than the analysis based on Fourier transform method .They can extract the fault features more effectively, particularly in a weak signal extraction. In this paper, the wavelet packet analysis theory was discussed in detail. The singular value which extracted using wavelet packet was used as the feature vectors. It demonstrated that the singular value can reflect the characteristics of rub-impact signal much better, and provide much sample data for the future of fault diagnosis.Besides characteristics of rubbing fault diagnosis and the basic theory of support vector machines were described systematically in this paper. The support vector machine was used to get the sample data classified, in order to achieve fault diagnosis and pattern recognition. The method proposed in this paper was compared to the neural network classification method to demonstrate it is a much more effective method. Finally a software system for simulation was developed, in which Mat lab and VC ++ were used together. The results show that the proposed intelligent fault diagnosis method is feasible, and have the broad using prospects in many areas.