A multi-constraint hosting capacity assessment framework for coupled power-transportation networks considering heterogeneous EV charging behaviors

Abstract

With rapid EV growth, urban road and distribution networks are increasingly coupled through charging demand. This paper proposes a hosting-capacity assessment framework for coupled power-transportation networks (CPTN) that combines heterogeneous vehicle-agent modelling, dynamic traffic assignment and time-series distribution-network power-flow analysis. Private EVs are represented by data-driven Markov trip chains with an end-of-day charging rule, while electric taxis are modelled as continuously operating agents driven by time-varying origin-destination demand. A bi-level iterative mechanism captures congestion, charging-station queuing and traffic-flow redistribution. Using an IEEE 33-node distribution feeder coupled with a 44-node urban road network, the framework scans EV penetration from 10% to 40% under voltage, volume-to-capacity and queuing constraints. The maximum safe EV penetration is 30%; at 35%, the minimum charging-node voltage falls below 0.95 p.u. and the scenario is judged as voltage-limit violation. Beyond identifying a single limiting penetration level, the framework also traces how traffic saturation, charging-service pressure and feeder-voltage margin evolve across penetration scenarios. The results show that the proposed heterogeneous and queue-aware representation can distinguish warning states from true constraint violations: road saturation appears before the final hosting boundary, whereas voltage becomes the binding constraint at 35% penetration. This provides a compact basis for planning-oriented CPTN assessment because it links vehicle operating patterns, station service capability and distribution-network security within one repeatable simulation workflow.