fault_detection_bearings

Diagnosis of Bearings Faults

Published: 28 November 2013

One of the most common sources of failures for the electrical machines is coming from the bearing faults and based on the related literature these faults are almost 40% of all the faults that electrical machines might have.

The major reasons for the bearing faults are mainly the thermal, electrical, and mechanical stresses, while these faults can be grouped into three types: Inner race fault, outer race fault and ball fault. The bearing faults cause increased temperature and excessive audible noise, reduced working accuracy, high vibration in the machines and produce harmonics of the stator current which lead to other types of faults such as air gap eccentricity.

Vibration analysis is one of the most widely utilized condition monitoring approaches for bearing faults but it requires additional sensors. On the other hand many techniques have been utilized for bearing fault detection and diagnosis such as the Stator Current Spectral Analysis, the Park's Vector Approach, Wavelets, the Hilbert transformation. Moreover, in the last years several Artificial Intelligence and Support Vector Machines (SVMs) have been also appeared in the related literature.

Overall, the aim of our group is to perform leading research in the field with extended experimental evaluations.

Contact: