A Spatial-temporal Charging Load Forecasting Modelling of Electric Vehicles Considering Urban Traffic Network

With development of electric vehicles (EVs), the grid is gradually facing additional pressure on peak load. Therefore, it is necessary to have accurate charging load forecasting as users behavior is full of uncertainty influenced by the traffic network and trip chains. In this paper, a novel charging load forecasting model is proposed to maintain the spatial-temporal characteristic of EVs considering different trip chains and traffic network, where Logit delay function is applied to calculate the road resistance. The normal distribution is used to fit the departure time of each stroke, and the improved Dijkstra algorithm is adopted to select the shortest time-consuming driving path. Finally, the efficiency of proposed model and method is assessed in the road network of test urban area.