Abstract
In order to solve the problem of fault feature extraction and selection of rolling bearing based on wavelet packet decomposition, we adopt wavelet packet energy and wavelet packet sample entropy as diagnosis features respectively, and diagnose bearing faults by using support vector machine (SVM). The influence of the level of wavelet packet decomposition, fault severity and load on the diagnosis accuracy of rolling bearing is analyzed. The results show that there is an optimal level for the feature extraction of a certain length of vibration signal based on wavelet packet decomposition. By comparison with the method using wavelet packet sample entropy, the method using wavelet packet energy is more stable with the working load variation and more suitable for the incipient fault diagnosis of rolling bearings.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have