Tri-Level Adaptive Robust Optimization for the Optimal Charging Coordination of EVs Considering Grid Impact

• Published in: IEEE access 2024, Vol.12, p.169755-169767
• Main Authors: Silva, Jessica Alice A., Camilo Lopez, Juan, Banol Arias, Nataly, Zenichi Terada, Lucas, Rider, Marcos J.
• Format: Article
• Language: English
• Subjects: Adaptation models; Adaptive control; Adaptive systems; Constraints; Coordination; Distributed generation; Electric power distribution; Electric vehicle charging; Electric vehicle fleets; Electric vehicles; Energy distribution; Linear programming; Mathematical programming; mathematical programming with equilibrium constraints; Nodes; optimal charging coordination; Optimization; Optimization models; Power control; Power dispatch; Reactive power; Reactive power control; Renewable energy sources; Resilience; Robust control; tri-level adaptive robust optimization; Uncertainty; Voltage control
• ISSN: 2169-3536; 2169-3536
• DOI: 10.1109/ACCESS.2024.3498600

The rapid integration of electric vehicles (EVs) into power grids introduces substantial operational challenges, particularly in managing the grid and ensuring efficient energy distribution. Addressing these issues, this paper presents a novel tri-level adaptive robust optimization model for EV charging coordination, explicitly considering active and reactive power control (RPC) and its impact on the distribution network. The first level optimizes the operation of unbalanced three-phase AC distribution systems through economic dispatch of distributed generation (DG) units, modeled using mathematical programming with equilibrium constraints (MPEC). The second level minimizes the energy non-supplied (ENS) to EVs during charging, considering both grid constraints and DG dispatches. The third level introduces an adaptive robust approach to handle uncertainties related to demand, renewable energy generation, and EV initial states of charge. This tri-level model, formulated as a min-max-min optimization, is solved using the column-and-constraint generation (C&CG) method. Validation on 25-node and 123-node systems equipped with dispatchable DG units, photovoltaic systems, and an EV fleet managed by a charging point operator (CPO) demonstrates the model's ability to mitigate uncertainties and balance conflicting interests between the CPO/aggregator and the distribution system operator (DSO). The results show that incorporating RPC reduces grid impact and ENS by up to 17.26% in the 25-node system and 30.68% in the 123-node system, highlighting the effectiveness of the proposed approach in enhancing grid resilience amidst increasing EV penetration. This work offers a comprehensive and scalable solution for private charging infrastructures, providing critical insights into improving grid resilience, optimizing EV charging operations, and effectively balancing the interests of both the CPO/aggregator and the DSO.