Research on aero-engine bearing fault using acoustic emission technique based on wavelet packet decomposition and support vector machine

The normal running of aero-engine is an important assurance for the plane's flight safety. The development trend of condition monitoring and fault diagnostic technology of engine is to be real-time and intelligent. In this paper, a new method based on the acoustic emission technology is proposed for the fault diagnosis of aero-engine bearings. Using the analytical method of wavelet packet decomposition and support vector machine to identify the fault type of bearings. Firstly, the measurement signal is decomposed by using wavelet packet decomposition. And the energy value of different frequency bands was extracted as eigenvector, which is used as training sample of support vector machine. The analysis result of the experimental data proves that the method has high recognition rate. This essay provides a new approach to the research of fault diagnostic of aero-engine bearings, and it throws new light on the further research.