Research On Identification And Suppression Of Low Frequency Oscillation Of Wind Power Integration Based On ICA And Wigner-ville Distribution

With the trend of large-scale interconnection of modern power system,ultra long distance and ultra high voltage line construction has become the first choice of power construction.And the country new energy projects have had good results,wind power is the most important.When large-scale wind power equipment is connected to the original system,the transient stability of power system presents new characteristics and new problems,and the low-frequency oscillation problem caused by small disturbance is being paid more attention by researchers.Therefore,in this thesis,the wind power system and grid-connected model are established by selecting DFIG as the object.According to the advantages and disadvantages of previous studies,a low frequency oscillation research method based on Wide Area Measurement System is proposed,which combines independent component analysis and Wigner-Ville distribution,and the appropriate suppression strategy is selected.The Power System Stabilizer is used to increase the damping of the system and to suppress the low frequency oscillation.First of all,the mechanical and electrical models and DFIG wind turbine models of wind turbines are established.Then through the analysis and discussion of the generation mechanism of the low frequency oscillation under the wind power grid,the negative damping mechanization is selected for this thesis.However,the low frequency oscillation signals monitored from the system have nonlinear characteristics.For dealing with the problems,the common research methods have different disadvantages,so this thesis proposes a new method to discriminate and analyze the low frequency oscillation.The constructed ideal signal and the power oscillation signal in the four-machine two-area system with wind turbine are processed by ICA,and the low-noise signal is identified by WVD.Through the example analysis and simulation comparison,the research method of low frequency oscillation based on ICA-WVD has better identification effect and higher identification accuracy than the conventional identification method,and its expression form is also more concise and clear.The error identification problem of Prony algorithm is avoided.The feasibility of the method is verified by the measured power data of a wind farm.It also lays a solid theoretical foundation and detailed scientific basis for the design of PSS suppression parameters.