Application Of Model Predictive Control In Reactive Voltage Regulation Of Wind Farms

Several factors such as the dependence of doubly-fed induction generators(DFIGs)on the grid,the volatility of wind power and the coordination of the DFIGs in wind farm have become intertwined and overlaid.The voltage control of wind farm at the point of common coupling(PCC)is facing severe challenges.This paper combines the advantages of model predictive control(MPC)in dealing with complex system models,uncertainties and control constraints,and analyzes and discusses the effects of various reactive compensation devices on voltage.Based on the principle of various reactive compensation devices and MPC theory,the MPC is applied to the voltage control of wind farm at PCC,realizing the wind farm process voltage control.Following shows the main work and achievements of this paper:For the scenario of the voltage of point of common coupling(PCC)is susceptible to the random fluctuation of wind power when the grid at the connection point is weak,this paper proposes a coordinated wind farm voltage control strategy based on hierarchical model predictive control.Firstly,in the adaptive adjustment layer,a voltage adaptive regulation strategy is proposed according to the reactive power capability of the wind farm and the voltage prediction track.Secondly,in the reactive power distribution layer,the reactive power demand capacity is optimized,and an improved proportional allocation strategy that can consider the reactive power margin of each DFIG is proposed.Finally,in the tracking control layer,the control error is corrected in real time according to the state prediction and reference as well as feedback information.Through hierarchical model predictive control,the prediction information of different timescales in each layer can be fully utilized,and the control between different layers can be coordinated effectively.To deal with the voltage variation caused by the random fluctuation of the wind power,this paper presents a two-layer voltage control strategy of wind farms based on stochastic model predictive control(SMPC).In the upper layer,on-load tap changer(OLTC)and capacitors are employed to regulate the large voltage variations with long timescales.In the lower layer,static var generator(SVG)and DFIGs are coordinated to timely eliminate the slight voltage variations with short timescales.Through hierarchical control,SMPC can be implemented separately in each layer to cope with fluctuations of different timescales,and continuous and discrete reactive power compensation devices are effectively cooperated between each layer.At the same time,a reactive power substituted strategy between SVG and DFIGs is proposed to keep the dynamic reactive power reserve of SVG maximally for the potential voltage fluctuation.To deal with the wind pool area's voltage dramatic variation caused by the random fluctuation of the wind power,this paper presents a distributed hierarchical voltage control strategy of wind farms based on stochastic model predictive control(DSMPC)along with the consensus-based distributed information.In the upper layer,OLTC and capacitors/inductors are centrally employed to regulate the large voltage variations with long timescales.In the lower layer,each wind farm constitutes an independent control area,and optimizes the reactive power output of each DFIG in the wind farm.The required information of each wind farm is interacted with its immediate neighbors,and computations with distributed optimization algorithm are carried out in the interconnected areas based on alternating direction method of multipliers(ADMM)to coordinate the reactive powers in different wind farms.A typical structure for centralized integration of 9 wind farms was used for the case study to validate the proposed control strategy under steady-state condition.Moreover,the robustness against a communication link failure and plug-and-play capability of the proposed voltage controller were tested.