Voltage fluctuations occur frequently in wind farms as a result of wind power perturbation

while reactive power control of various devices is always uncoordinated and often has a time lag when the system operates in the traditional control mode that only uses the current data for decision-making. In this case

a method for coordinated control of the reactive power of wind turbine generators (WTGs) and static var generators is proposed based on model predictive control (MPC). Unlike the traditional voltage control method

the control method proposed has the realization of optimization of the voltage control curve and the dynamic process of reactive power regulation in the future window of time as its objective. An auto regression moving average (ARMA) model is used to predict the active power of WTGs while the reactive power of WTGs and the SVG is forecasted using a first-order inertia link and a proportional integral (PI) link

respectively. Voltages in the wind farm are solved with the help of a sensitivity matrix. Thus an optimization model is developed for both the minimization of PCC voltage deviation and the maximization of SVG reactive power reservation in a future window of time. The optimizing model is solved using CPLEX tools with the dual simplex algorithm. In order to test the performance of the MPC controller

multiple cases with different wind circumstances are simulated on a DIgSILENT platform. Simulation results reveal that by predicting the states of various instruments in wind farms during a period of time in the future

the proposed control method is able to rationally arrange fast/slow reactive power to respond to predictable changes of the system in advance and keep voltage stable. This work is supported by National Key Basic Research Program of China (973 Program) (No. 2013CB228201)

National Natural Science Foundation of China (No. 51277105) and National Science Fund for Distinguished Young Scholars (No. 51025725).