A bi-level optimal dispatching model for EV-based virtual energy storage system

Owing to shifts in global energy construction, the use of electric vehicles (EVs) has increased rapidly. In order to promote the consumption of renewable energy and eliminate the potential adverse effects of high EV penetration, this paper proposes the novel concept of the virtual energy storage system (VESS) and a corresponding bi-level optimal dispatching model. The VESS consists of all EV batteries currently connected to the grid in the same moment. The installation location of the VESS depends on the distribution of available charging piles, while the VESS' capacity is variable depending on EV parking times and quantity of electric batteries. The upper level of the associated bi-level VESS dispatching model determines the VESS' scheduling strategy, and a complex time-sequential trip chain model considering spatiotemporal EV distribution and road congestion level is proposed to obtain the available VESS capacity in the distribution system. The model's lower level decides the specific power allocation for each EV. Data from National Household Travel Survey 2017 (NHTS2017) and EV database are employed to mimic the VESS in the distribution network. A case study comparing the EV uncontrolled charging load and optimal scheduling demonstrates that the proposed VESS dispatch strategy can consume surplus photovoltaic power and balance load fluctuation.