Research On Optimal Operation Mode Of EHV Power Grid Considering Wind Power

With the increase of global energy demand and the increasingly prominent environmental problems,wind energy has been gradually valued and favored by all countries in the world.As the most promising clean energy for economic development in renewable energy,the wind power Grid-connected capacity has increased year by year.However,China's energy and load are obviously unbalanced in the regional distribution.A large number of wind power needs to be transferred to the load center by means of EHV power grid for remote absorption,which poses a serious challenge to the economic operation of EHV power grid.In the actual management of EHV power grid enterprises,the only alternative way to reduce losses is to change the output of direct dispatching plant.However,the diversity of wind power generation scenarios makes it a difficult problem to formulate a reasonable output plan of plant.From this point of view,this paper will conduct the research from the following aspects:Firstly,the probability model of wind speed and the power characteristic model of wind farm are analyzed.In view of the wind speed data that do not conform to a certain probability model,the non-parametric nuclear density estimation and three-parameter Weibull distribution model are used to fit the wind speed probability model.Then the fitting results of the two models are evaluated by chi-square test,which verifies the hypothesis of the wind speed probability model.At the same time,the wind farm output model is established and the corresponding wind power output scenes are generated by substituting the wind speed data.Secondly,a wind farm output scenario analysis method based on Latin hypercube sampling and improved K-means clustering is proposed.Latin hypercube sampling technology is used to sample the probability distribution model of wind speed and generate a large number of wind power output sampling scenarios.Then K wind power output operation scenarios are generated by improved K-means clustering method.The optimal clustering number is determined by clustering validity index.The rationality and validity of scenario analysis method are verified by practical examples.Furthermore,the actual Northeast China Power Grid is analyzed and studied based on PSASP.After dividing the Northeast Wind Power Station and Direct Direction Plant into zones,the changing trend of network loss is studied when the output of wind power and plant changes.Based on the breadth-first search method,an optimal operation scheme to minimize the network loss of the system is obtained,and the corresponding operation and dispatching suggestions are put forward.Finally,an extra-high voltage power grid optimization model with the objective ofminimizing network loss is established.And an improved particle swarm optimization(IPSO)algorithm suitable for complex practical examples is proposed to solve the model.The particle position is initialized by Sobol sequence sampling,and Levy flight is introduced into the particle global search mechanism to improve the search divergence and ergodicity.Taking the actual 500 kV grid of Northeast China Power Grid as an example,the superiority of the improved algorithm is verified.At the same time,the optimization results under different scenarios are compared and analyzed,which provides some references and suggestions for the optimal operation of the actual EHV system with wind power.