Detecção de defeitos em motores de indução utilizando a análise das correntes de linha

Abstract

In industrial environments a great challenge is to keep driving machines in order to deliver the maximum capacity with minimal resources, as these are extremely important for diferent processes. It is essential to keep the high availability of them, but there are many variables involved in its operation, in which the induction motor has a significant participation. Thus, there is a great requirement regarding the reliability of production processes and reducing costs, in this scenario the predictive maintenance techniques become a essential tool. In this work was evaluated the EPVA technique (Extended Park Vector Approach) for the detection of motor defects in a noninvasively ways, avoiding the need to stop the production process for the verification of an abnormal condition. The technique allows machines online monitoring projects reading only two electrical currents with the installation of Hall-effect current sensors in motor power panels allied with specific software. As proposed, was verified in a satisfactory way that the technique through the Concordia vector analysis and its corresponding frequency spectrum, detect voltage imbalances defects, misalignment between machines and the opening of the rotor bars, considering two load situations, proving the theory studied.