In order to improve the monitoring accuracy of faults which consist of DFIG stator and rotor winding fault and bearing fault, different fault characteristic frequency is obtained by the Simulink fault model of DFIG, the accuracy of the fault model is verified by theoretical analysis on the fault characteristic frequency.Using ZoomFFT and multi window Fourier transform to extract the fault characteristic quantity, carries on the frequency spectrum analysis and the contrast.The experimental results show that the rife-vincent window FFT have a higher accuracy, resolution and reliabilityin the treatment of DFIG fault of anti-jamming; the fault characteristic frequency is influenced by the the change of wind speed , in order to achieve a better effect in the double fed motor state monitoring ,the fault reference value should be adjusted by the actual wind speed.