Economic real-time energy trading services for electric vehicles with uncertain mobility demand

Real-time energy trading services have recently been proposed to facilitate the transition from vehicles with combustion engines to e-vehicles. Such services involve e-vehicle owners and a provider of energy trading skills, where the latter manages e-vehicle (dis-)charging at real-time energy prices. We propose and evaluate a data-driven optimization approach for the operational management of services that address e-vehicle owners with uncertain mobility demand and with a solar generator. We formulate the management problem as a sequential decision problem, and propose to solve it by integrating a lookahead policy with a safety energy buffering approach. We study by computer simulation the economic sustainability and the practical viability of the services under different degrees of mobility demand uncertainty, and for different e-vehicle owner types. We assess the negative impact of the uncertainty on both financial value and mobility demand satisfaction of the services. We show that our optimization approach with time-dependent energy buffering enables high mobility demand satisfaction while achieving, even in the case of extreme mobility demand uncertainty, on average at least 92% of the financial value that can be achieved under quasi-deterministic mobility demand. We show that combining our approach with departure time flexibility is sufficient to reach perfect mobility demand satisfaction and a further increase of the financial value.