Answers are grounded in this paper and cite their sources.
This paper proposes a stochastic chance-constrained bilevel optimization framework for electric vehicle charging station (EVCS) integration and reserve provision, using a single-bus equivalent of a smart distribution system for coordinated day-ahead energy and reserve-capacity scheduling. The upper level minimizes the distribution system operator’s (DSO) operating cost, while the lower level minimizes the parking lot operator’s (PLO) own net operating cost by scheduling EV-fleet charging, discharging, and reserve-capacity provision under uncertainty in response to given economic signals from the DSO. The framework jointly considers renewable-generation forecast errors, uncertain load demand, and the time-varying availability of a heterogeneous electric vehicle (EV) fleet. Renewable forecast errors contribute to reserve requirements, while normally distributed arrival and departure times determine hourly connection probabilities for each vehicle category. The aggregation of vehicle connection states within each category motivates a Gaussian approximation of category availability. Conservative availability is calculated for each vehicle category at the prescribed confidence level and aggregated to derive time-varying EVCS power and energy limits and the reserve margin associated with EV availability. A load demand chance constraint yields the corresponding risk-adjusted demand and associated load uncertainty margin. The operating-cost objective also accounts for battery degradation. The bilevel model is transformed into a mathematical program with equilibrium constraints (MPEC) using the Karush–Kuhn–Tucker (KKT) conditions and subsequently reformulated as a mixed-integer linear program (MILP) through Big-M linearization. The framework is evaluated over a 24 h horizon on three test systems comprising 3 to 80 distributed generators and 260 to 1100 EVs. Across the examined confidence levels, EVCS reserve provision reduces operating cost by 1.7 to 7.7%, while higher confidence levels increase scheduling costs.
No takes yet. Share an insight, caveat, or question.
Koufalexis et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: