Broken rotor bar detection using stator current data in induction machines

Induction motors are the most widely used electrical motors in industrial applications due to their reliability, relatively high efficiency, and lower maintenance demand. Any fault taking place in induction motors may result in long down-time and lead to financial losses. Therefore the reliability of induction motors are quite important for mainly industrial and other applications. Broken rotor bars are the most common fault types that may be seen in induction motors. This thesis presents a combination of NLMS (Normalized Least Mean Square) filter and Hilbert Envelope Analysis methods for detecting broken rotor bars of the squirrel-cage induction motors. The proposed method is tested by using experimental data. Stator current data is used as experimental data. The experimental data is produced by using three identical rotors (one healthy rotor, one with three broken bars, and one with five broken rotor bars) in an induction motor under different loading (25%, 50%, and 100 %) conditions. The results of the recommended method are compared to the results of Fourier Transform. The results shown that the proposed method can be successfully used for detecting the rotor faults of squirrel-cage induction motors.