Electric Vehicle Charging Load Prediction Considering Multi-source Information Real-time Interaction and User Regret Psychology

Abstract

This paper proposes a method for predicting the charging load of electric vehicles considering real-time interaction of multi-source information and user regret psychology. Firstly, the starting and ending points of the private car and taxi trips are obtained through the travel chain theory and the OD matrix method respectively, and the driving routes are planned using the Dijistra algorithm; Then, a practical speed-flow model based on the real-time traffic statistics of the road network is constructed to calculate the real-time vehicle speeds in each road section of the road network. A power consumption model per unit mile for the electric vehicles is constructed considering the environmental temperatures and vehicle speeds, and the power consumption is calculated. Next, the electric vehicle charging station selection model based on the regret theory is proposed, taking into account the factors such as the charging tariff, driving time, queuing time and power consumption along the routes. Based on the information from the multiple sources such as the traffic network, vehicles, fast-charging stations, the distribution network and so on, a multi-source real-time interactive EV charging load prediction system is established. Finally, the Monte Carlo method is used to simulate the travel and charging processes of the private cars and taxis, and the spatial and temporal distribution of the charging loads in the region is obtained. Simulations are carried out on a regional traffic road network and a typical distribution network and the effectiveness of the proposed charging load forecasting method is verified. The simulation results show that the timely information interaction of multiple sources will have an impact on the spatio-temporal distribution of the charging loads.