Modeling and Forecasting Spatial Charging Demand for Electric Vehicles: Comparative Analysis of Agent-Based, Statistical, and Geospatial Approaches

This paper presents a comparative analysis of three primary approaches to modeling spatial demand for electric vehicle (EV) charging: geospatial, statistical, and agent-based. The calculated spatial EV charging demand results are verified using real-world data from public charging station usage in a major Russian city over a period of 2.5 years. A detailed description of the three models under consideration is provided, and the calculation of the most promising locations for charging station deployment using each model is performed. The comparative analysis of the simulation results is based on metrics for the accuracy of calculating EV charging demand and for determining the ranking of the most attractive locations for EV charging station placement. Based on the analysis results, the…