Electric Vehicle Charging Load Forecasting Considering Urban Land Use Types and Road Network Information

Spatial-temporal prediction of electric vehicle (EV) charging load is an important prerequisite for optimal scheduling and planning of charging stations. Aiming at the limitations of the coupling analysis of road network traffic flow and urban functional areas, which makes it difficult to accurately analyze the dynamic stochastic characteristics of charging load under high penetration rate, this paper proposes a road network model that combines urban land use types to predict EV charging load. First, the city area was functionally divided by ArcGIS, and the topology was constructed based on the road network characteristics. Second, a single EV mobility model that has been refined in terms of both travel characteristics and charging characteristics. Finally, the Monte Carlo method is used to simulate the charging behavior of EV in a region of Lanzhou City and to compare and analyze it with the fixed power consumption parameter model. The results show that the overall load accuracy in the study area is improved by 9.97% compared with the traditional model, and the evening peak load accuracy is improved by 2.53%. The feasibility and validity of the proposed model are verified, which lays the foundation for the next step to study the cooperative operation of road-vehicle-network with different attributes.