Lumping and electromechanical transient modeling of large-scale wind farm

With the increasing capacity of wind farm merged into the grid, their impact on the power system is also growing. It needs to establish an accurate and reasonable wind farm model. Large wind farms usually consist of hundreds or even thousands of wind turbine generators (WTGs). The amount of computation will be huge if each WTG is carried out a detailed model. So an equivalent model for wind farm is needed: wind farm is classified based on the specific situations, then lumped into a number of units. This paper adopts hierarchical cluster procedures to classify the wind farm, makes hierarchical clustering analysis for all WTGs and divides them into several clusters by taking transient voltage characteristics of each WTG as clustering target. Lumped process which takes the actual power characteristics of the cluster as a reference value adopts parameter identification. Previous WTG models mainly based on electromagnetic transient modeling, containing many power electronic devices is complex and has poor convergence and slow computing speed. So it is not suitable for large-scale grid-connected analysis. This paper establishes electromechanical transient model of doubly fed induction generator (DFIG) in the MATLAB/Simulink environment based on vector control strategies. Considering that the dynamic process of converters and switch states changed fast (usually microseconds) in the electromagnetic transient simulation, while the process of electromechanical transient is slower (usually milliseconds), the electromagnetic transient process of converters and switch can be ignored because it has ended when electromechanical transient simulation begins. Finally, electromechanical model parameters are identified based on the measured data. This paper identifies the model parameters combined with the Northwest Power Grid. The simulation result accurately shows that the power output of electromechanical transient simulation is very close to the measured data when the wind speed a- d voltage fluctuates and it verifies the reliability of the lumped method and electromechanical model.