A feedback-integrated framework for resilient and distributed scheduling of electric vehicles under uncertain charging characteristics

Abstract

Emerging innovation in smart charging for plug-in electric vehicles (EVs) has the potential to achieve significant economic benefits. In several works, smart charging encourages the use of EVs as a flexible resource by modifying their power consumption through a demand response (DR) program. However, it is promptly assumed that EVs are always responsive and accept the smart charging signals with no fault. In practice, due to uncertainties such as random EV mobility, volatile battery charging characteristics or charging component failures, some EVs would be unable to accept the assigned charging signals dispatched from a central server. Therefore, this article proposes a feedback loop to predict EV charging behaviours and thereby adaptively tune the time-based control signals dispatched to individual EVs. Moreover, a parallel-operating distributed DR algorithm is proposed which aims optimal EV scheduling under charging uncertainties while reducing the need of private information sharing. The proposed distributed algorithm allows increased EV user privacy, fast convergence properties and optimal operation under communication disruptions and delays. The effectiveness of the proposed methods are also numerically exhibited for varying penetration of EVs within a low-voltage (LV) distribution test network.