Homogeneity-based approach for bearing fault detection in induction motors by means of vibrations

Electrical machines, in particular induction motors (IM), are important parts in an industrial plant, representing an 89% of power consumption. Bearings are important parts of the induction motors and one of the principal causes of their malfunction; hence, bearing fault early detection is very important, however its detection is a challenging because the measured signals are acquired in noisy conditions and have transient characteristics. Hence, a system to detect the potential faults into bearings of rotatory machinery in their early stage can have a potential benefit in industry. In this work, a novel proposal that makes use of the homogeneity (HO) algorithm for the bearing defect, in particular the outer race (OBD), detection is presented. The HO method is introduced for the first time to detect the changes produced in the normal regime (steady-state) vibration signals of an IM by the OBD. These signals can contain subtle modifications on motor dynamic features due to the fault presence. The presented results show the proposed methodology is capable of distinguishing between a motor with OBD and a healthy motor with a high efficiency.