Classification of Roller Bearing Faults using Iterative Classifier

Bearings play an important role in power transmission systems and to ensure healthy functioning to get the best output. Therefore, a frequent and reliable inspection scheme needs to be employed to intervene before any critical damage occurs. Vibration studies, conducted on bearings reveal useful information, which helps distinguish different working conditions. The data recorded is used to calculate statistical values, which can help diagnose faults in any bearings. This study focuses on classifying the conditions of bearings using statistical features with the help of the Iterative Classifier technique coupled with Multilayer Perceptron in accurately detecting a faulty bearing.