Aiming at the problem of fault detection of high voltage disconntector mechanism in traction power supplysystem

a fault diagnosis method based on stator current characteristics and SVM is proposed. By analyzing the rela-tionship between the torque fluctuation and the stator current

the functional relationship between the stator currentfluctuation and the effective value of the side frequency band and the fundamental wave is established

and it is con-cluded that the side frequency component and the effective value of the fundamental wave of the stator current can ef-fectively show the type of mechanism fault. Then

the combination of low-pass Butterworth filter and digital FIR filteris designed. The amplitude frequency characteristic value of the 45 ～ 55 Hz side frequency and the fundamental ef-fective value time characteristic value of the opening and closing process are obtained by analyzing and processingthe sampled stator current data. These two current characteristics are used as the input of SVM to build the SVM faultdiagnosis model. Finally

the experimental results show that the overall accuracy of the model is 97.8%.