Artificial neural network in the application of the doubly-fed type wind power generator parameter identification

Doubly-fed wind power generator (DFIG) type is the mainstream model of the wind turbine at home and abroad. To study the impact of large-scale wind power grid on power system reliability, it must have accurate model parameter. In the Matlab/Simulink environment, We have set up a simulation model for the wind turbine grid and have got the measured data. The artificial neural network algorithm is applied to the “gray box” model and it has the effective network output function curve fitting. Therefore, it chooses the alpha beta coordinate system mathematical model of the generator. In the case of model recognition, using artificial neural network algorithm adopts step by step identification strategy, and it can get stator self inductance, mutual inductance, mutual inductance between rotor and stator. It subjects for the study of large-scale wind power grid to provide reliable theory.