Wind Power Converter Open Circuit Fault Diagnosis Method Based On The Current

At present,with the large number of countries in the world to develop and implement the development of wind power program makes the production of wind turbines continue to increase,which make the system fault type and location to determine the follow-up maintenance become the focus of reliability of the study.In the existing wind power generation system,the direct-drive wind power system is more widely used because of its simplicity,robustness,and low construction and maintenance costs.Because the geographical location of wind power generation system is more special,and wind power system is a huge and complex system,any one of the parts of the abnormal or minor failures may lead to the occurrence of serious damage to the entire system or even a direct shutdown,the converters as the core of the entire wind turbine functional components,due to stability and poor working conditions and other reasons,leading to the whole system is running directly connected with the converter.Therefore,it is of great practical significance to carry out the research on the fault diagnosis and detection of wind power converter.Based on the analysis of the basic structure and working principle of the direct-drive permanent magnet wind power generation system,the mathematical model of the machine-side and rotor-side converters in three different coordinate systems is analyzed.And the simulation model of direct drive type wind power generation system is built by using Matlab2014b/Simulink software platform to simulate and obtain the three-phase current output signal of the converter under different fault conditions,which lays the foundation for the fault diagnosis of converter.Through the analysis of the main types of common open circuit faults and the influence on the whole wind power system,the changes of the three-phase current waveform in different open-circuit faults are compared.In this paper,the basic theory of variational modal decomposition is described,and the construction and solution of variational problem are introduced.A three-phase current signal of machine-side converter is chosen as the research variable,and the method of decomposing and extracting effective features base variational mode is proposed.The kurtosis and mean value of the IMF components are calculated as the eigenvalues and the eigenvector is calculated.The simulation results show that the eigenvector has good state separability for different open faults and has good diagnostic and recognition effect on the fault of the converter.The support vector machine is used to idendify the open circuit fault type of wind power converter.The parameters of SVM classifier are optimized by cross validation method;select the sample training set and test set,and normalize the sample eigenvalue to train the SVM model;the test set samples were used to test the SVM and the results of different fault types were obtained.The experiment results verify that the method based on variational modal decomposition and support vector machine can improve the open fault diagnosis accuracy of wind power converter.