An Evolving Approach to Fault Detection of Rolling Element Bearings

This work proposes use a methodology for fault detection of Rolling Element Bearings using an online, evolving approach. The proposed methodology is composed of feature selection, statistical feature extraction, fault detection processing, low pass filter, and a time filter. The central part of our approach concerns the processing itself, in which we use an online learning evolving algorithm, named TEDA for fault detection. The proposal has been validated using data from bearing faults with different diameters. The results show the precision, recall, and f1 score of the different test scenarios performed.