Bearing Fault Diagnosis Based on Cyclic Statistics Method

In order to reduce the influences caused by background noise and interference and exact the fault characteristic frequency, the basic theory of second-order cyclic statistics is studied in this chapter. The excellent demodulation ability of second-order cyclic statistics is proved by simulative analysis. In the bench test, the fault characteristics frequency of bearing outer raceway and its harmonic can be recognized clearly in frequency domain through the spectral correlation density when cyclic frequency is zero. The result has higher signal to noise ratio (SNR). Compared with the traditional spectrum analytical methods, the impacts of background noise and interference are furthest reduced by cyclic statistics and the fault characteristic frequency of bearing is identified accurately.