Fault severity diagnosis of squirrel-cage induction motors in transient regime based on curve fitting

Abstract

Many motor current signature analysis (MCSA) methods have been proposed for the rotor fault diagnosis of induction motor. However, few method provides the quantitative measurement. In this paper, a method is proposed to diagnose rotor fault severity based on the transient current. First, wavelet transform is used to analyze the stator current and detect rotor bars. Then, a curve fitting algorithm is proposed to diagnose the asymmetric rotor condition and give the precaution value and alarm value. Lastly, a fault severity factor is presented based on the two aforementioned values. The proposed method is demonstrated by simulation experiment, which shows that it can give the evaluation of rotor asymmetry and fault severity.