Spatio-temporal Multi-dimensional Load Forecasting Of Electric Vehicles And Its Schedulable Potential Capacity Assessment Method

Electric vehicle charging load prediction is of great importance to the safe and reliable operation of the distribution network,and a high penetration charging load can provide considerable schedulable potential capacity to support grid frequency and peak shaving.Therefore,it is an important issue in the development of intelligent distribution networks to study two methods of charging load forecasting in a single time dimension and a complex spacetime dimension,and a method for evaluating the potential capacity of electric vehicles.A single time dimension charging load prediction method classifies cars,establishes a mathematical model of time characteristic variables through data analysis,derives the charging probability at each moment,and uses the Monte Carlo simulation method to obtain the car charging load.This method uses the daily driving mileage obtained by data analysis to replace the subjectively given initial state of charge(SOC)to derive the charging duration,and uses a more random charging probability to replace the calculated charging period to determine the charging load.Taking a city as an example,the daily load curve of a car is predicted and compared with the results of common load forecasting methods.The results show that this method can predict the charging load simply and quickly.The complex space-time dimension charging load forecasting method constructs a twodimensional dynamic traffic behavior model supported by space-time characteristic variables through data analysis.Then,in order to explore the characteristics of continuous charging and centralized charging of vehicles,two charging situations of continuous charging and centralized charging were set to construct a charging behavior model.Finally,the car traffic-charging behavior is simulated by the Monte Carlo method,the car load prediction results at different times and regions are calculated,and the influence of the load on the power grid is analyzed.The results show that this method can reasonably describe the characteristics of the twodimensional uncertainty of the vehicle's space-time,and can more accurately predict the charging load.The method for assessing the potential capacity of electric vehicles is based on analyzing the traffic-charging behavior of the vehicle and then evaluating the potential capacity.First,an accurate travel model is established through data analysis of car travel data,and then the vehicle's state of charge(SOC)curve is obtained by simulating the traffic behavior and charging behavior of the car,and finally considering the differentiated demand on both sides of the gridusers,according to the grid and user scheduling Constraint calculation of car dispatchable potential capacity.In order to study the dispatching potential under different dispatching methods,this paper sets three dispatching types under the constraints of the three types of frequency modulation methods in the power grid,and obtains the dispatching potential in four different regions and the impact of dispatching capacity on the power grid,and conducts other methods.Compared.The results show that the method in this paper can scientifically evaluate the dispatchable potential capacity of electric vehicles,and can provide reference and basis for grid dispatching.