Grid Rolling Optimization Operation Of Wind Power And Electric Vehicles Based On Distributed Algorithm

The consumption of fossil energy has brought about environmental pollution and energy depletion.As a result,traditional power generation is gradually being replaced by new energy power generation.While new energy power generation is responsible for power generation tasks,it also participates in power generation scheduling.At the same time,with the advantages of electric vehicles(EVs)and the government and EV-related departments encouraging the purchase of EVs,and issuing a number of preferential policies for EV charging,EVs have penetrated the power grid on a large scale,and energy is shared with the power grid.However,the randomness of new energy power generation and EVs will bring greater challenges to the dispatch of the power grid.The power of new energy generation is difficult to accurately predict,the amount of information communication when EV is charging,and EV owners may suffer privacy leaks.In view of this,a distributed framework with the power supply side of the power grid,EV load aggregates and EV owners as the main body is established in this paper.On the basis of this framework,an improved Alternating Direction Multiplier Method(ADMM)is used to solve the rolling optimization scheduling model,so as to achieve the goal of satisfying the interests of multiple entities in the power grid.First,a BP neural network model based on error correction is used to predict wind power,and the prediction results of this model and a single BP neural network model under the same sample are compared,and it is found that this model has higher prediction accuracy.At the same time,in order to consider the fluctuation of wind power prediction errors,the distribution characteristics of the errors generated during the prediction using the established prediction model are analyzed.Under the condition of given confidence,the fluctuation range of the predicted value of wind power is obtained to provide basic data for the subsequent wind power dispatching.Secondly,the charging model of a single EV and the differentiated models that lead to different EV travel rules due to the geographic location of different areas and the nature of the car owner’s work are analyzed.The process of rolling dynamic dispatch is introduced,and a rolling optimization model is established for the purpose of smoothing grid load fluctuations,increasing wind power consumption,reducing operating costs of EV operators,and improving charging satisfaction of EV owners.Then,a distributed real-time optimization framework is constructed,which divides the grid into layers:distribution network operators and local wind farms,EV aggregates in different regions,conventional loads,and a large number of EVs.Based on the distributed framework,the improved ADMM and the improved Grey Wolf Algorithm(GWO)are used to solve the established rolling optimization model.The improved ADMM only needs to transfer information between two agents when iterating,and does not require multiple agents to do it at the same time,so it can achieve the parallel solution of the optimization goals of each agent.Finally,the actual data in a certain area is used for simulation analysis.Before and after optimization,load variance,wind power consumption,EV operator operating costs,and EV owner’s charging satisfaction were compared,and the effectiveness and convergence of the ADMM method were improved,as well as forecast information,grid electricity prices and owner preference coefficients on the dispatch results are analyzed.The simulation results of the calculation example show that the established model can achieve the purpose of orderly charging and absorbing wind power for EVs while taking into account the interests of all parties in the power grid.