Fractal dimension theory-based approach for bearing fault detection in induction motors

Abstract

Induction motors, vital elements into the industry, are more likely to be influenced by different faults during their lifetime service. Even when they can keep working without affecting the line processes, in most cases, an increase in the production costs usually occurs. Bearing fault detection is an important topic due to the fact that this failure yields an increase in both vibration and temperature, among others, which can produce in other systems joined to the induction motor similar issues. In this regard, a monitoring system capable of detecting bearing fault in the induction motor condition is desirable in industry. In this work, a new methodology based on fractal dimension theory, a concept from the chaos theory, for outer race bearing defect (OBD) detection is presented. The fractal dimension (FD) theory is introduced for the detection of anomalies produced by OBD in the steady-state vibration signal of an induction motor, since this signal might have subtle changes on its dynamic characteristics due to the fault. The obtained results show that, as expected, the measured signal has the assumed changes, leading to have a methodology with a higher overall efficiency for distinguishing the fault and the heathy states.