Abstract
A fault feature extraction model based on the PCA and wavelet packet is proposed to describe the characteristics of gearbox fault feature, which is expressed by low amplitude of the vibration signals, and easy to be disturbed by system and noise. Firstly, the PCA is used to reduce the correlation between the data dimension and the data. Then, the gearbox signals are decomposed by wavelet packet, and reconstructed based on the frequency bandwidth characteristics. After choosing those main frequency band which reflects the change of signal caused by the fault, and normalizing the selected frequency band, then the fault characteristic value is obtained. Finally, the vibration signal of the gearbox is treated as an example to verify the effectiveness of the method. The comparative analysis shows that the combination of PCA and wavelet packet is more effective than the wavelet packet.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have