In order to extract the fault feature of wind power converter accurately and steadily, a fault diagnosis method is proposed in this paper, which is based on variational mode decomposition (VMD) and support vector machine (SVM). Firstly, grid-side three-phase currents and rotor-side three-phase currents are collected and decomposed into a number of intrinsic mode functions (IMFs) through VMD. In the next step, skewness value and kurtosis value are calculated as feature vectors for each IMF and these feather vectors are selected as the input of SVM classifier. Finally, the diagnosis results will be obtained by classification results. In addition, the simulation results reveal that the method proposed is effective in diagnosing circuit faults of wind power converter.
展开▼