| In recent years,wind power technology has made great progress and development.With the increasing scale of wind farm,the number of wind turbine units in wind farm is increasing.If each unit in the field is modeled in detail,it may lead to pay a long time to simulate and calculate,and also bring about difficult convergence of power flow calculation.Therefore,it is of great significance to simplify or equivalent wind farm model to wind farm modeling.In this thesis,on basis of mastering the principle of the front-end speed governing wind turbine,the overall structure of the unit is analyzed.The model of the unit is established and the output characteristics of the unit are analyzed.A one-dimensional-nonlinear-wake model is derived for the unreasonable assumption of the variation trend,it is different with the wake diameter of the Jensen wake.In the study of the model,the variation of turbulence intensity,wind shear effect and thrust coefficient with the working wind speed of the wind turbine is fully taken into account,which shows that the wake diameter is a parabola.By compared with the measured wind speed curves,it is shown that the one-dimensional-nonlinear-wake model can accurately simulate the trend of wind speed.For the external characteristic method of wind farm modeling,the measured wind speed and active power of each wind farm unit is analyzed in this thesis by means of mathematical statistics method.It is obtained the measured wind speed power characteristic curve of wind turbine unit.The mapping relationship between blade tip velocity ratio and wind energy utilization coefficient is obtained by the curve,and the wind turbine parameters which can reflect the actual output characteristics of wind turbine are obtained,and the wind turbine model is established.The measured wind speed data of each wind turbine in a wind farm are analyzed.According to the magnitude and fluctuation of the wind speed data,the improved maximum tree method is used to realize the wind turbine cluster,and the equivalent model of the wind farm is established.The simulation results show that the equivalent model based on the measured wind power curve is basically consistent with the measured output power curve,and in the case of wind speed fluctuation and system failure.It can effectively describe the output characteristics of wind farm and improve the accuracy of wind farm equivalent modeling.In view of the uncertainty of equivalent model caused by cluster results in wind farm,the wind farm aggregation modeling method based on multi-machine representation method is adopted in wind farm modeling.The probability transfer matrix is established according to the Markov property of wind speed,and the wind speed is taken as the cluster index,and the two-step method is used to classify wind turbine aggregation.According to the Fisher criterion,the clustering results are analyzed.The aggregation model of wind farm is obtained by taking the highest probability of each quarter as the final clustering result.Compared with the output power curve of the detailed wind farm model,the simulation results show that this method is effective in modeling wind farm with front-end speed regulating wind turbine,and the model has high precision. |
