Forecasting Charging Load Of Large-scale Electric Vehicle

The superiority of electric vehicles becomes more prominence on improving environmental quality,stemming climate warming and relieving the energy crisis.Developing electric vehicles has been treated as a long-term national development strategy in our country.As the progress of technology and the promote of policies,electric vehicles will achieve large-scale development.If not deployment and management in coordinated,large-scale charging demand is bound to cause great influence to power grid.In order to meet the huge demand of the charging preferable,at the same time alleviate the grid influence of the electric vehicle charging.Forecasting load growth caused by electric vehicle and mastering the spatial and temporal distribution of EV charging load have a certain practical significance to provide decision-making basis for power grid upgrading and economic operation.The ownership of electric vehicle has a decisive role to the general trend of charging load.This paper embarks from the diffusion regularity of durable goods,a combination forecasting model minimizing the error sum of squares of predicting electric vehicle ownership on the basis of the Logistic and Gompertz extension prediction model is proposed.The method overcomes the limitation and one-sidedness of single forecasting model,improving the prediction result.Predicting the ownership of electric vehicles in our country,the effectiveness of the model is verified.The analysis of the influence factors is a prerequisite for forecasting charging demand.This paper analyze the influence factors of charging load respectively from the electric vehicle,charging facilities,the user's travel characteristics,charging management and other aspects.According to influence factors,analyzing and modeling of different kinds of vehicles'travel rules and charging behavior.At the same time,different charging behavior of the electric vehicle can be divided into certain charging behavior and uncertain charging behavior,using the fuzzy inference to get the charging probability of uncertain charging behavior in different destinations.In order to get the size and spatial distribution of the electric vehicle charging load in predicting area.First of all,conducting spatial-division according to the nature of land use and distribution of charging infrastructure,dividing predict area divided into different function blocks,street blocks and areas.This paper will treat feasible charge places in the city as the main research object to conducted the charging load prediction,such as the residential,the workplace,the commercial blocks,charging station and the bus parking lot.Using the Monte Carlo simulation method to calculate different kinds of charging place charging load,obtain time distribution of charging load in different function blocks.On the basis of considering load coincidence factor,we obtain street blocks' and areas' charging load.Based on the analysis of two layers,we can get in the prediction of spatial and temporal distribution of the electric vehicle charging load.At last,we take a certain area of A city as an example,forecasting its future charging load,obtained spatial and temporal distribution of charging load in the area.In addition,the effects on the power grid in different charging place is analyzed.To verify the proposed method can be more specific and effective to provide certain reference basis for the expansion of the power distribution network planning and coordination scheduling.