Study On The Multi-Machine Equivalent Model Of Wind Farm And Its Application

With the aggravation of the energy crisis and the environmental pollution, power generation based on clean energy has become the inevitable trend of the future power grid. Wind power is one of the fastest growing new energy power generation technologies in the world. It is also the most mature technology too. It has developed rapidly during recent decades. However, the security problems of power grid caused by wind power has occurred occasionally. With the increasing scale of the wind farm, the impact of the wind farm on the power system will be further intensified. Simulations are a common way to analyze the power system with large wind farm. But it is difficult to establish the detailed simulation model of the wind farm. Therefore, it is urgent to study the equivalent model of wind farm to analyze the influence of large-scale wind power interconnected with grid.In order to study the equivalent model of wind farm, the simulation model of a single doubly-fed induction generator(DFIG) is established firstly. The normal operaion characteristics and transient characteristics of DFIG are analyzed by simulations. The fluctuated wind and the changing of the referenced reactive power are applied to testify the dynamic response of the DFIG. The characteristics of low voltage ride through(LVRT) with different control strategies and Crowbar are also analyzed.Then, the multi-machine equivalent modeling method of wind farm based on fuzzy C means(FCM) clustering is proposed in this thesis based on the monitored states of DFIG. The output active power, the output reactive power, the RMS value of voltage and current of DFIG are selected as the grouping indices. According to the given number of equivalent machines, the grouping problem of wind farm is transformed into a clustering problem. The fuzzy membership function of wind generator in a group is discussed. The objective function of FCM clustering algorithm is established. By solving the optimal clustering center and fuzzy membership matrix using iteration, the clustering results are obtained. According to the clustering result, the equivalent model of each groups is deduced. Multi-machine equivalent model of the wind farm is obtained.Finally, the proposed method is applied to analyze the Ningxia power grid operation. Ningxia grid is a multi-send HVDC system with wind plants. This is very complicated. For the summer big operation mode in 2014 and the winter big operation mode in 2016 of Ningxia power grid, the equivalent model of Mahuangshan wind farm is established. The output characteristics of wind farm with different number equivalent machines are compared by simulations. The simulation model of Ningxia power grid is established with PSCAD. The simulation results are compared with the power flow calculation results of PSASP. The comparison verifies the steady state validity of the proposed method.The proposed method can reduce the computation of the equivalent modeling of wind farm after the calculation scale and the number of equivalent machines are specified. The effectiveness of the equivalent modelling is improve with the guarantee of the accuracy.