Behaviour Modelling and V2G Scheduling of Large-Scale Electric Vehicles Using Dynamic Optimal Power Flow

With rapid proliferation of electric vehicles (EVs), their uncoordinated charging processes being mismatched with renewable generation poses significant challenges to electrical grid stability. This paper proposes a hierarchical vehicle-to-grid (V2G) scheduling method to optimise the aggregated import and export behaviour of large-scale EVs for enhanced grid flexibility subject to individual charging requirements. The travel behaviour of a limited number of monitored EVs is first characterised by modelling distributions of arrival time, parking durations and initial state of charge levels. The modelled characteristics of monitored EVs are then extrapolated to a given number of large-scale EVs by the Monte Carlo simulation. The EVs are aggregated and regarded as an energy storage system connected to the local node within a distribution network. The aggregated available power and energy of EVs along with the minimum energy import required prior to next trips are estimated at each time step across a day, and then introduced into dynamic optimal power flow to schedule V2G interaction. Simulation results demonstrate that the proposed method effectively schedules charging and discharging of EVs to absorb otherwise curtailed renewables and replace part of conventional generation during peak hours, respectively, while respecting power- and energy-related operational constraints of EVs.