Vehicle-Originating-Signals for Real-Time Charging Control of Electric Vehicle Fleets

We propose the vehicle-originating-signals (VOS) approach for charging control of a fleet of electric vehicles (EVs) in an electricity distribution network (DN). The goal of the approach is to manage the EV load such that the aggregate power consumption, including inflexible demand and solar generation, closely follows a given target profile. The VOS approach enables EVs in a fleet to compute signals reflecting their need for charge and willingness to supply power. An aggregator collects these signals and implements the control with minor computational effort. We evaluate the VOS approach on a scenario for Munich, Germany, based on a mobility survey and real electricity demand and solar generation data. First, we compare our approach against a centralized optimization in terms of objective fulfillment and solving time. We show that it achieves a competitive performance, especially for large vehicle fleets. Second, we present a statistical method to evaluate the performance and limitations of our approach. We identify a recommended range of the vehicle fleet size for a given load magnitude and show that the target profiles can be met with a defined certainty. Finally, we present a method for further reducing the communication overhead with minor effect on performance.