V2G Regulation in Distribution Networks With Decision-Dependent Uncertainties: A Chance-Constrained Learning Method

Le résumé fourni par la source

With the proliferation of electric vehicles (EVs), coordinating Vehicle-to-Grid (V2G) technology in distributed networks can enhance grid flexibility. However, the uncertainties of EVs pose big challenges to the implementation of V2G. Especially in dispatch scenarios with real-time prices, the uncertain charging demand of EVs is impacted by real-time pricing decisions, which is a typical case of decision-dependent uncertainty (DDU). Moreover, the mapping between pricing decisions and the distribution of charging demand is ambiguous and challenging to characterize in practice. To address these challenges, this paper proposes a bilevel game framework considering DDUs for V2G regulation in a distribution network. A chance-constrained learning (CCL) method is proposed to capture the probability distributions of DDUs under different pricing decisions. This method relaxes the restrictive assumption in existing research that requires explicitly constructing the mapping between decisions and DDUs in advance. Instead, it directly constructs this mapping by machine learning, which gives more accurate modeling results for DDUs. Due to the incorporation of the neural network model into the optimization constraints by CCL, the original problem is transformed into a mixed-integer quadratic constrained problem (MIQCP) with bilinear terms. Then, an improved segmentation-based approximation method is introduced to reformulate the original problem as an MIQCP with linear terms, which makes the problem solvable. Finally, the IEEE 123-bus and 500-bus systems are utilized to validate the performance of the proposed V2G regulation framework in terms of DDU modeling accuracy, system economics and stability.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.