Research On Distribution Network Planning Based On EV Uncertainty

Due to the global environmental problems and energy crisis caused by the widespread use of traditional vehicles,new energy EVs(EVs)gradually replace traditional vehicles as new vehicles,and the continuous increase in the scale of EV access and new energy penetration has led to the transformation of energy strategies.Not only need to consider the uncertainty of EV network access,but also need to develop a matching distribution network planning strategy.Therefore,it is necessary to study the uncertainty planning of EV access network in the distribution network environment.This paper starts from the impact of EV on the distribution network and builds an improved Bass EV ownership forecasting model.The main research contents are as follows:Firstly,the research background and objective significance of EV access distribution network planning uncertainty are introduced.The effects of EV access distribution network,EV load forecasting methods,and EV access distribution network uncertainty planning methods are reviewed.In view of the uncertainty of the EV market shape,based on the traditional Bass car ownership forecasting model,the constant parameters(maximum potential m)in the traditional Bass innovation diffusion theory cannot track the defects in the number of potential EV users.Build a model for predicting the EV inventory of Bass with the maximum potential change model,and simulate it through the matlab2018 a numerical fitting toolbox.A comparative study shows that the improved EV inventory prediction model of Bass greatly improves the fitting accuracy,thereby ensuring EV.Accuracy of inventory forecasts.Secondly,in view of the phenomenon of low randomness and intermittentness caused by the strong randomness of the charging load sequence,based on the characteristics of the bidirectional long-short-term memory network that is good at processing time series data,a EV charging load prediction based on Bi-LSTM network Method,the model can rely on the ability to train sequence features based on the positive and negative bidirectional existence of time series,can accurately capture the intrinsic characteristic information of EV charging load,and then construct the EV load of(Bi-LSTM)bidirectional long-term and short-term neural network The prediction model is compared with the prediction effect of the EV charging load model of LSTM,and the prediction results of different methods are compared.The result proves that the EV charging load model based on Bi-LSTM has high prediction accuracy,which provides the location and capacity of EV charging stations.A reliable basis.Finally,based on the distribution network data of EVs in a certain district of Shanghai,combined with the relevant knowledge of fuzzy mathematics theory,the multi-objective optimization problem is transformed into a single-objective optimization problem that takes into account the satisfaction of decision makers.And EV charging load substation and grid planning plan,and formulating a distribution network planning plan with EV in a certain area.The results of EV access distribution grid planning are given.The rationality of the mathematical model is verified by application examples.And scientific,put forward a new theoretical method for future EV access distribution network planning.