Energy trading for electric vehicle aggregator under random participation of EV owners

This paper proposes a novel two-stage stochastic program to facilitate two-way energy trading (including up-regulation service) for an electric vehicle aggregator (EVA) in the electricity markets. Different from the existing works, the proposed model considers the randomness in the charging and discharging preferences of the EV owners on the operating day. Such uncertainty arises because the EV owners may randomly decide to participate on the operating day. Further, the EVA faces other uncertainties related to both the market-level data (i.e., electricity prices, real-time dispatch signal) and the fleet-level data (e.g., energy consumption of EVs, arrival and departure times) prior to participating in the electricity markets. The computation complexity of the proposed problem is tackled by constructing a small set of scenarios and applying the Benders decomposition algorithm. The performance of the proposed model is analyzed by comparing with the traditional model which ignores the randomness in the participation of the EV owners. The results show that the proposed model significantly outperforms the traditional model in terms of the total expected net cost incurred by the EVA. Further, we analyze the market-wise and fleet-wise performance under both the models through extensive simulations.