Fault feature extraction of rotor broken bar and stator winding inter-turn short circuit of asynchronous motors

In order to accurately identify the stator and rotor faults of three-phase squirrel cage induction motor, it is necessary to extract the fault characteristics of the motor accurately and quickly. A normal induction motor, the stator phase current only have the fundamental component in theory, but it will produce harmonic components of different frequencies or change the energy of the original harmonic components when a fault occurs. Convert the fault current to frequency domain and get its spectrum, will be found that the amplitude of fundamental current is very large and the amplitude of harmonic current is very small, and their frequency is very close when the rotor bar is broken. Because of the influence of fundamental current, it is difficult to identify the fault feature components in the spectrum. In this paper use a hybrid genetic algorithm based on Nelder-Mead simplex method to estimate the fundamental current, obtain the expression of fundamental wave current and remove it from the fault current. In this way, the fault characteristics of motor stator or rotor can be obtained.