LAUSR

The three-phase induction motor (TIM) is present in most industrial processes, being the main responsible for the electromechanical conversion in the sector. The total breakdown of these machines can compromise the production of entire sectors in a manufacturing environment, and hence, the need for early fault detection. In this context, broken bar failures are the most common on the TIM rotor. This article analyzes the chaotic behavior of one stator current of a TIM, with different levels of broken rotor bar failures. The characterization of the failure is performed by analyzing the behavior of the chaotic variable called density of maxima. The proposed approach has advantages over classic motor current signature analysis techniques in that it does not require long acquisition windows (in this work, it was possible to perform detection using a range of 0.35 s) and enables fault detection even for motors operating under variable load or with low slip. The technique also has a lower computational cost compared with classical detection methods. The obtained results indicate that the method is also able to identify different severity levels with the TIM under nominal load.