Low-Carbon Economy Optimization of Park-Level Integrated Energy Systems Considering Temporal and Spatial Scheduling of Large-Scale Electric Vehicles

Park-level integrated energy systems (PIESs) realize coordinated carbon reduction via multi-energy coupling and multi-agent flexibility, yet few studies jointly optimize the temporal and spatial scheduling of large fleets of electric vehicles (EVs) for low-carbon economic dispatch. This paper develops a bi-layer scheduling model for PIESs integrated with large EV fleets. The upper layer adopts multi-objective optimization to minimize operating costs and maximize carbon abatement, coordinating CHP, boilers, energy storage, PV and aggregate EV power across a 24 h horizon. A carbon emission intensity index (CEI) is defined to trace embodied carbon flows and impose system carbon constraints. The lower layer minimizes distribution network power loss by rationally assigning EV charging/discharging loads to network nodes subject to voltage, line capacity and charging pile limits. NSGA-II with fuzzy decision-making solves the upper multi-objective MINLP, and Gurobi addresses the lower MILP power flow model. Simulations on a modified IEEE 33-node park system with four comparative cases validate the proposed strategy. Results show that, compared with disordered EV charging, the bi-layer scheme cuts total cost by 18.2%, net carbon emissions by 42.1% and network losses by 17.4%. Sensitivity analysis proves that the carbon price and EV penetration remarkably affect economic and low-carbon performance. This two-stage spatio-temporal optimization framework provides an effective tool to harness the low-carbon flexibility potential of large-scale EV fleets in industrial and commercial parks.