Spatiotemporal Prediction of Urban EV Charging Demand Through Synthesizing Electromobility Trajectories in an Integrated EV–Transportation–Land Use–Energy Framework

In the absence of empirical data on urban electric vehicle (EV) mobility and charging demand, this paper presents a methodology to predict spatiotemporal urban charging demand by synthesizing electromobility trajectories (e-trajectories) within an integrated EV-Transportation-Land Use-Energy (EVTLUE) framework. The EVTLUE comprises: (1) a power-based EV model for estimating energy consumption; (2) an integrated graph-based street network and land-use polygons; (3) a dynamic traffic module utilizing Google Maps traffic data; and (4) a planning model for optimally deploying slow chargers in land-use and fast-charging stations on high-traffic streets. For each EV trip, an e-trajectory is synthesized based on the interactions between the EV user and EVTLUE across driving, charging, and idle sub-trajectories. The driving sub-trajectory is simulated using user driving behavior to infer trip purpose, destination, and route choices, and to generate profiles of speed, energy consumption, and battery state. The charging sub-trajectory is derived from user charging behavior, modeling decisions regarding the time, location, and mode of charging under four strategies: obligatory, convenient, postponement, and delayed charging. Daily user e-trajectories are synthesized to predict individual daily charging demand and the spatiotemporal distribution of slow and fast charging demand across urban areas. A numerical simulation for 45,360 EVs in Campinas, Brazil, reveals a total urban charging load of 1,164.18 MWh, with 77.66% attributed to slow charging and 22.34% to fast charging. Slow charging demand (904 MWh) is distributed across homes, workplaces, and public locations with respective shares of 42.21%, 44.16%, and 13.63%. Fast charging demand (260.1 MWh) is met by 71 optimally planned FCSs.