Spatial-temporal Distribution Prediction of Charging Load for Electric Vehicle based on Dynamic Traffic Flow

Abstract

On the basis of velocity-flow-density relationship and traffic-energy consumption relationship, this paper proposes a prediction method of the spatial and temporal characteristic of electric vehicle charging load using the traffic data. By analyzing the residential travel data, a probability model was built to generate trip chains of a day, which contain destination and start time. Then vehicle transfer model was used to simulate driving vehicles on the roads and SOC could be calculated by the road condition and temperature. Drivers would charge vehicles when SOC is below the charging threshold. Finally, by using the Monte Carlo method, charging load of a real traffic model was calculated according to charging demand from all electric vehicles at different time and location during the area.

Electric vehicle charging load prediction considering user behavior characteristic (Withdrawal Notice)

“…Literature [4] considered the parking laws and driving and charging behaviors of large-scale EV in different time and space, and used Monte Carlo method to predict the temporal and spatial distribution characteristics of EV charging load, but did not consider road types and congestion conditions. In literature [5][6][7][8][9] , the influence of road network information on EV driving law is considered, and the random dynamic characteristics of EV are simulated by using travel chain or traffic start-stop method to achieve the prediction of EV charging load. In literature [10] , the influence of traffic and temperature on power consumption per unit mile was comprehensively considered, and the randomness of paths was simulated by Markov decision process to obtain EV charging load.…”

“…On the driving path, the SOC of EV is calculated according to Equation (6), and the corresponding functional area is determined to charge. If charging is required, update EV SOC according to Equation (7) and record its corresponding charging amount; Otherwise, continue until you complete the journey chain.…”

Research on dynamic temporal and spatial distribution characteristics of electrical power system voltage based on random forest algorithm

“…Affected by the disturbance, the variation difference in average voltage fluctuation rate at different monitoring points can roughly describe the power distance and fluctuation rate from the fault area. 8 Based on this, the average fluctuation rate of voltage drop can be taken as one of the indexes in the selection the quantitative indexes of the voltage's dynamic spatial and temporal distribution. t 0 and t min respectively represent the time of starting voltage drop and the time of reaching the minimum voltage under the condition of faulty electrical power system.…”

Electric Vehicle Charging Load Demand Forecasting Model based on Spatial and Temporal Characteristics

“…Chen Lidan considered factors such as temperature and traffic road conditions to predict the charging load of EVs under different penetration rates and different scenarios [6]. Song Yunong established a probabilistic model for the spatial and temporal distribution of EV load based on dynamic traffic flow by combining a traffic road network model, a vehicle spatial and temporal transfer model, and a residential travel probability model [7]. Zhang Meixia established a hierarchical charging decision model considering user charging variability by analyzing parking hour adequacy and time-sharing tariff [8].…”