Hierarchical Reactive Power Optimization Control Strategy Of Wind Farm Cluster Considering Power Prediction

With the carbon neutrality and carbon peak proposed,many countries around the world call for a greater effort to develop and build new energy sources.As an important part of new energy,it will be the key development direction of the future power system to integrate large-scale wind power into the grid.Compared with thermal power,hydropower and other traditional power sources,wind power is more volatile and intermittent.Therefore,connecting the large-scale wind farm cluster to the power system frequently causes unbalanced distribution of power flow,which will threaten the safety and stability of the power system.Considering the wind farm cluster has reactive power compensation capability,wind turbines(WTs)can output reactive power to play an active role in voltage regulation when there is a voltage fluctuation near the wind farm cluster.The voltage control structure of the wind farm cluster can be divided into grid layer,inter-field layer and inner-field layer.In existing reactive power and voltage control strategies of the wind farm cluster,there are shortcomings such as large amount of calculation data,uncertainty of wind power,and difficult coordination between different control layers.Therefore,the further research is carried out and the specific work content is as follows:(1)The dynamic reactive power demand assessment method for the voltage weak area wind power integrated is proposed.Firstly,the random matrix theory is applied to identify weak voltage nodes and areas in power system where the large-scale wind farm cluster is integrated.Secondly,it is judged whether wind power needs to play the role of voltage regulation according to the identification result.If the area where the wind farm cluster is integrated is the weak voltage area,the short circuit fault is set for the weak voltage node in the area adjacent to the basic operating status of the power system,and the transient voltage stability index is calculated.Thirdly,a BP neural network is trained to fit the mapping relationship between the operating state of the power system and the transient voltage stability index.Finally,based on the sensitivity method,the reactive power increment amount of the wind farm cluster is calculated quickly and quantitatively online,which improves the setting speed and accuracy of reactive power demand.(2)The voltage droop control strategy of WTs considering the uncertainty of wind power is proposed.Firstly,a model of ultra-short-term wind speed prediction based on wavelet neural network is established and the uncertainty of active power output from WTs is analyzed by wind power characteristic curve.Secondly,a voltage droop control strategy for WTs is proposed based on the mixed Gaussian probability distribution.Then the reactive power capacity of WTs can be calculated according to the predicted output of active power,and the probability density and expected value of voltage droop gain can be calculated based on the method of adaptive voltage droop and mixed Gaussian probability density estimation.Finally,the expected value of voltage droop gain is applied to the control link of WTs,and reactive power output of WTs is allocated reasonably.(3)Combined with the reactive power demand assessment method of the wind farm cluster and the voltage droop control strategy of WTs,hierarchical reactive power optimal control strategy for the wind farm cluster considering power prediction is proposed.Firstly,voltage distribution characteristics of wind farms are analyzed and the expression of reactive power loss of transmission lines between wind farms is deduced.Secondly,the voltage control structure of the wind farm cluster and the main functions of each control layer are described.Thirdly,based on the evaluation results of reactive power demand of power grid layer,a calculation method of voltage droop coefficient base value considering the reactive power loss of transmission line is proposed and the calculation result is solved by PSO algorithm.Fourthly,the prediction values of voltage droop expected gain of WTs is calculated based on the voltage control strategy in inner-field layer.Finally,IEEE New England test system containing wind farm clusters is built to verify the effectiveness of the proposed strategy in developing the voltage regulation capability of wind farms and improving the transient voltage stability of power system.