Rolling element bearing fault feature extraction using an optimal chirplet

Fault feature extraction from vibration signals is an important topic for fault diagnosis in rolling element bearings. However, the vibration signals measured from rolling element bearings are usually complex, and impulse components are usually embedded in strong background noise. In this paper, a novel method using an optimal chirplet with hybrid particle swarm optimization is proposed. The inner product absolute value of the vibration signal and the chirplet basis function is used as the fitness function. By heuristically searching the optimal parameters of the chirplet basis function, the optimal chirplet is further improved to increase its analysis results. The proposed method is applied to analyze vibration signals collected from rolling element bearings, and the results confirm that the proposed method is more effective in extracting fault features from strong noise background than traditional methods.