Online slight inter-turn short-circuit fault diagnosis using the distortion ratio of load current in a low-voltage induction motor

Abstract

In recent decades, the use of induction motors in industrial environments has increased, and the demand for both maintenance management and condition monitoring has expanded. A literature survey shows that the inter-turn short-circuit insulation failure of stator windings is one of the most likely faults to occur in motor-drive systems. This short-circuit fault is mainly due to the damage or deterioration of electrical insulation. Moreover, the short-circuit failure of one turn will initiate stator winding insulation failure, leading to the breakdown of the entire system. Hence, identification of one-turn insulation failures in early stages is essential. Thus, the present paper deals with such failures of the induction motor. First, the frequency spectrum of the load current is analyzed by fast Fourier transform, and the characteristic frequency components are extracted. Next, the distortion ratio of the load current is calculated using the above characteristic frequency components. Finally, a new diagnostic method applying a support vector machine is proposed, and its advantages are described. The experimental and diagnostic results are presented to validate the proposed analytical procedure using the distortion ratio.