Research On Equivalent Modeling And Parameter Identification Of Wind Farm

In recent years,the scale of wind farms have gradually increased.If each unit of large wind farm is modeled in detail,it will increase the amount of calculations and consume a lot of time.Therefore,it is necessary to conduct in-depth research on the equivalent modeling method of wind farms.With the complexity of power grids,fault conditions often lead to insufficient reactive power,resulting in a decrease in system voltage and reactive power.Static reactive power generators(SVG)are increasingly used in power grids and wind farms to compensate for reactive power and stabilize voltage.However,in actual situations,SVG often cannot provide accurate parameters,and most of them use typical parameters in simulation.So,the controller parameters need to be accurately identified.The content of this article is developed from these two directions,and the main work is as follows:Firstly,the permanent magnet direct drive wind turbine model in BPA is briefly introduced in this paper,which includes the power control model under normal conditions,the power control model during LVRT,the power control model after the end of LVRT,the active and reactive current limit model and full-power converter model.These models provide a theoretical basis for subsequent simulations.Secondly,the methods of clustering wind turbines in wind farms are summarized.Considering the LVRT process and terminal voltage of the units,it is proposed that the first grouping is carried out on the basis of conduction of the unloading circuit,and for the units whose unloading circuit is not acted,they are divided into two groups according to the steady-state voltage and transient voltage sag values.After grouping,we calculate the parameters of equivalent model,and then calculate the error between equivalent model and detailed model under multiple conditions.By comparing with the single-machine equivalent method,a conclusion is drawn that the error of the equivalent model calculated by the method proposed in this article is greatly reduced,which proves the effectiveness of this equivalence method.Then we check the LVRT capability of the equivalent model,and calculate the conditions that the equivalent model needs to meet under various working conditions.Then,according to the output of different wind turbines,different failures,and different failure times,the LVRT capability of the equivalent model under various working conditions is verified.Then,a method of SVG parameter identification based on fault response curve and sensitivity analysis is proposed.Firstly,the parameters are screened according to the sensitivity of the trajectory of each parameter.And then through a large number of simulations,it is known how do each parameter of SVG affect the fault response curve.We select different segmented curves and use the grid search method to initially identify the parameters to narrow the parameter range.Finally,the particle swarm algorithm is used to accurately identify the SVG parameters,according to the comparison results of traditional parameter identification method and SVG parameter segmentation identification method based on fault characteristics,the method in this paper is proved to be effective.In the end of the paper,we design a software platform that can realize equivalent modeling of wind farms,verification of wind farm LVRT capability,and SVG parameter identification functions.Through this platform,as long as we have prepared the model files needed for simulation,the simulation calculation can be finished through some simple operations on the interface.This software greatly reduces the complexity of work,and it has been applied in Zhejiang Electric Power Research Institute.