Spectral kurtosis based on evolutionary digital filter in the application of rolling element bearing fault diagnosis

Abstract

Rolling element bearings are essential components in rotating machinery. It is important to detect the bearing fault as earlier as possible. It is known that spectral kurtosis (SK) is sensitive to impulse signal and has been widely used to detect bearing fault. Whereas, the incipient fault of bearing is weak and difficult to extract especially in a complex rotating system. Focusing on this issue, this study proposed a hybrid approach using evolutionary digital filter (EDF) and SK to detect rolling element bearing fault feature. Firstly, the signal to noise ratio of the raw signal was enhanced by EDF in a self-learning process. Then, the optimal band was detected using fast SK. Envelop analysis is later employed to extract the fault characteristic frequencies. The proposed approach was verified by numerical simulation and experimental analysis. Results show that the proposed SK-based EDF yields a good accuracy in bearing fault diagnosis.