Search for: All records

With the increasing integration of electric vehicles (EVs), there is a growing recognition of the contribution of EVs to the flexibility of distribution networks. However, quantitative methods for precisely assessing the flexibility that EVs can provide to the upstream power grid have not been fully explored. To fill the research gap, this article proposes an online hierarchical aggregate flexibility characterization method, which considers the distinct roles of the distribution system operator (DSO) and charging station operators (CSOs). Under this framework, each CSO first characterizes the aggregate EV power flexibility of its charging station as a time-decoupled flexibility region. We prove that any trajectory within the region can be disaggregated into feasible individual EV charging strategies. Upon receiving the aggregate EV power flexibility regions submitted by all CSOs, the DSO solves an optimization problem to derive the aggregate power flexibility region of the distribution network, and we also prove that any power level within this region can be disaggregated into a feasible solution satisfying the network constraints. Furthermore, to account for the impact of the latest dispatch order on future EV power flexibility, real-time feedback is designed and integrated into the hierarchical framework. Numerical experiments on a multiphase distribution network with 123 buses demonstrate the effectiveness and advantages of the proposed method. It is worth noting that the proposed online method outperforms the offline benchmark with perfect information, due to the consideration of dispatch signals. 

This article proposed a weighted Bayesian potential game (WBPG)-based architecture framework to achieve mutual benefits for the noncooperative coordination between electric vehicle charging stations (EVCSs) and distribution system operators (DSOs), ensuring mutual operational benefits under demand forecast uncertainties and heterogeneous EVCS types. The proposed framework enables EVCS to dynamically optimize their charging strategies and bidding behaviors, thereby maximizing individual utilities while improving global welfare at the Bayesian Nash equilibrium (BNE) as a win-win outcome for EVCS and DSO. To accommodate the specific information-gathering and processing constraints of each EVCS, this article employs fictitious play (FP) and Cournot adjustment within the competitive process among EVCS as well as a dynamic pricing mechanism, enabling each EVCS to assess its expected utility and engage in self-optimizing bidding in a Bayesian context. The proposed framework also provides a systematic analysis of spot price volatility driven by EV charging demand uncertainties. Finally, the proposed approach is validated through comprehensive numerical simulations in both a small-scale scenario (15 EVCS) and a large-scale real-world case (600 EVCS). The results demonstrate that the WBPG framework outperforms traditional mixed-integer linear programming (MILP) optimization and noncooperative Stackelberg game models, achieving an 11.4%–19.4% increase in global welfare with moderate computational time. These findings underscore the advantages of WBPG in enhancing the resilience and efficiency of EVCS-DSO coordination.