Research On Power Flow Optimization With Wind Power By VSC-HVDC

The large-scale grid-connected wind power and the increasing annual installed capacity of wind power have brought new challenges to the operation of power systems,especially with the development of Voltage Source Converter based HVDC(VSC-HVDC),an increasing number of wind farms are connected to the VSC-HVDC,thus forming a VSC-HVDC large power grid.Therefore,it is urgent to model and analyze the VSC-HVDC power grid with wind farms,but the power flow optimization itself is a multi-variable,multi-constrained,non-linear large-scale planning problem.How to solve the optimization power flow model of VSC-HVDC connected to wind farms has also become a problem.In order to solve the problems mentioned above,this thesis mainly deals with the following aspects:Firstly,the wind power correlation model based on Copula function is established.The entropy weight method is used to evaluate various Copula function models by weighting various evaluation indexes,and the function model which best reflects the correlation of wind power samples is selected.The actual wind power data of a certain wind farm is used to verify the example.Secondly,a trend-based optimization model for VSC-HVDC power grid based on scene method is established.The latin hypercube sampling is performed on the established wind power correlation model based on Copula function,and several power values considering correlation are obtained.The K-means clustering method under the guidance of density clustering is used to cluster multiple power values.The obtained finite power value can well reflect the actual situation of wind farm output correlation.The wind power uncertainty is transformed into the determined wind power by the scene method,and the wind farm is built into the model by flexible DC transmission,and the reliability simulation analysis of the scene method application is carried out from the two aspects of economy and voltage quality.The simulation analysis studies the effects of wind power correlation,wind power grid connection mode and converter station control mode on power flow optimization.Thirdly,an improved teaching and learning optimization algorithm is proposed,which is applied to the optimization of the VSC-HVDC power flow.While researching the teaching and learning optimization algorithm,the adaptive factor and the elite strategy are used to improve the algorithm in the two stages of "teaching" and "learning" from the perspective of economy and voltage quality.The results are compared with the particle swarm optimization method as well as traditional teaching and learning optimization algorithm,the feasibility of the improved algorithm proposed in this thesis is verified.Fourthly,the problem of unit combination is taken into account in the VSC-HVDC power grid,and the dynamic VSC-HVDC power flow optimization model of the wind-fire combination is formed.A generalized Benders decomposition method is introduced to divide the problem into the main unit combination problem,the wind power scene switching subproblem and the VSC-HVDC power flow safety constraint sub-problem.The three problems iterate with each other to finally determine the start-stop status and active output of the unit.The modeling method described in this thesis can well reflect the VSC-HVDC grid connected to the wind farm.The proposed teaching and learning algorithm has a good solution to such models,ensuring the safe and stable operation of the VSC-HVDC grid.