Integration of electric vehicle user charging preferences into Vehicle-to-Grid aggregator controls

Abstract

Vehicle-to-Grid (V2G) describes the concept of using the battery packs of parked electric vehicles (EVs) as distributed battery energy storage to support the power grid. The bi-directional power flow in a V2G network is managed by a so-called aggregator which has to consider the needs of both the power grid and connected EVs. ‘Range anxiety’ is a major factor holding back the EV market as many drivers consider range limitations as a major disadvantage of EVs over more traditional vehicles with internal combustion engines. Implementing V2G could potentially worsen this perception as connected EVs could actually lose charge over short periods when providing grid services. Enabling EV owners/users to state charging preferences via simple user interfaces might increase the acceptance of V2G and reduce range anxiety.The inclusion of user charging preferences puts further constraints on an aggregator and complicates charge/discharge scheduling. Taking such preferences into account and weighting them against (potentially conflicting) needs of the power grid is a major challenge for V2G scheduling and adds computational costs through increased complexity. However, depending on its application, the responsiveness of a V2G network is crucial and the scheduling process is hence time-sensitive. In this research, a method of capturing user charging preferences and accounting for the resulting constraints is presented. The impact of added complexity on the execution time of the scheduling process is addressed by resolving user inputs prior to the actual scheduling. The ability of each EV to either charge or discharge is scored taking into account user inputs and the current state of the battery pack. This scoring system informs the selection of suitable EVs before the scheduling process.