Vibration Characteristic Analysis and Feature Extraction of Bearing Coupling Fault based on Sparse Representation

Abstract

Sparse representation has a wide range of applications in the field of image processing and audio processing. Applying the sparse representation theory to the field of vibration signal processing can efficiently represent the periodic components of the signal. Through the Adams system simulation and measured signal rotor coupling fault data sparse representation， the vibration characteristics of rotor unbalance bearing coupling fault and the law of interaction between faults are studied under the sparse representation. The results show that bearing faults are usually submerged under coupled faults， and it exhibits weak fault characteristics compared to unbalanced faults. Aiming at the problem that the bearing fault feature is difficult to extract in the rotor unbalance bearing coupling fault， using the sparse representation method based on Gabor atom to match the periodic vibration component in the rotor unbalance bearing coupled fault vibration signal， and the spectral kurtosis algorithm is used to find the frequency band where the impact signal is located for bearing fault feature extraction. The effectiveness of the proposed method is verified by analyzing multiple sets of measured signals.

Keywords

• Sparse representation;OMP algorithm;Fault diagnosis;Vibration signal processing;Coupling fault