A Novel Bearing Fault Diagnosis Method Based on LMD and Wavelet Packet Energy Entropy

Abstract

AbstractRolling element bearings faults may lead to fatal breakdown of machines. Therefore, it is significant to be study bearings diagnosis, and the vibration-based methods have received intensive study because vibration signals collected from bearings carry rich information on machine health conditions, and it is possible to obtain vitalcharacteristic information from the vibration signals through using signal processing techniques. This paper proposes a novel vibration-based diagnosis method about bearing faults, first, a new pattern recognition method is proposed to diagnose bearing faults through using the interval value of the spectral peak frequency in the frequency domain; second, vibration signals of different parts faults of the bearings will be processed by different algorithm for precisely extracting the fault characteristics; and third, in order to extract transient characteristics from a noisy signal, the filter need to be developed and to further improve the signal-to-noise ratio (SNR), band pass filter is designed based on the PSD of vibration signals in this paper. The vibration signals collected from rolling element bearings are used to demonstrate the performance of the proposed method, andthe results verify the effectiveness of the method in extracting fault characteristics and diagnosing faults of rolling element bearings.