Spatiotemporal distribution prediction of electric vehicle (EV) charging load at highway service areas

As electric vehicle (EV) adoption increases, highway fast-charging demand becomes increasingly time-varying and spatially heterogeneous, highlighting the need for reliable demand estimates to support infrastructure planning and subsequent energy‑system studies. This study proposes an integrated vehicle-level Monte Carlo framework for estimating potential EV charging demand at highway service areas. A simplified graph-theory-based representation of a typical Chinese highway corridor is parameterized using representative network and traffic inputs. The framework integrates an origin–destination (OD) matrix, stochastic EV characteristics, travel routing, a physics-based energy‑consumption model, and SOC-based two-stage charging decisions. To capture the stochastic variability in EV travel and charging behaviors, the complete one-day simulation is independently repeated. The scenario-based results reveal distinct spatial and temporal variations in simulated charging demand across highway service areas. These findings provide preliminary references for identifying high-demand service areas, supporting charging infrastructure planning, and informing subsequent transportation‑energy system studies.