An improved algorithm for detection of rotor faults in squirrel cage induction motors based on a new fault indicator

Abstract

In this paper an improved fault detection algorithm is proposed to ameliorate the reliability of the rotor fault detection task. The proposed algorithm uses the Motor Current Signature Analysis (MCSA); it is based on monitoring the Relative Harmonic Indexes (RHI) as a new fault indicator. The most sensitive harmonics to the occurrence of rotor bar faults contribute to the calculation process of the RHI, which represents its main advantage compared to the classical fault indicators. The proposed algorithm requires only the stator current signal as input. For each data acquisition, it estimates the slip, identifies the frequencies and the amplitudes of the searched harmonics then it computes the RHI. After that, the algorithm normalises and classifies the RHI. The obtained results are displayed on the monitor screen of the personal computer. For any data acquisition, the proposed algorithm allows the user to know the motor state, the fault severity and the slip. A lot of experimental tests, carried out on a 3kW squirrel cage induction motor, confirm the effectiveness of the proposed algorithm to detect rotor bar fault under different operation conditions even at very low loads.