Application of Positive and Negative Separation Method in Stator Winding Fault Diagnosis

Present methods used in stator wingding fault diagnosis have many disadvantages, such as instability, complex calculation. To solve the problems, positive and negative separation method is applied in stator winding fault diagnosis. Firstly, acquired data passes low-pass filter to weaken the influence of high-harmonics and yawp, then it applies positive and negative separation method to extract fault feature. Power transformation can eliminate source judgment of negative part; Park transformation converts signal from 3-D coordinate axis to 2-D coordinate axis; Hilbert transformation has frequency phase-change quality to apart positive and negative part. Simulation results show that this method can avoid other un-fault factors´ effect and obtain positive and negative parts of stator voltage and current. It adopts stator negative current and negative impedance to describe fault severity degree. This method solves the difficulty in fault extraction, so it is useful to detect early fault of stator winding.