Monitoring and predicting charging behaviour for electric vehicles

In this paper the tracking and analytical infrastructure necessary to understand and manage the power demands of an electric vehicle fleet is considered. The data from a 230 day trial of 15 electric vehicles has been used to create a model day for the equivalent of 3000 vehicles in the North East of England road network. Basic analytical approaches are considered and possible future avenues are addressed. A general model for predicting vehicle charging is proposed. The comparative charging rates between morning and evening and the spatial distribution of the charging are all considered. It is found that evening charging is the most common period of charging but morning charging, due to its spatial density, poses the most risk for local power. In more general terms the use of individual vehicle tracking systems is found to be an ideal system for determining the current and future state of power consumption for electric vehicles. Having such infrastructure is important from both a local area management perspective and also as a tool to inform decision makers and politicians on the future needs for electric vehicles. As the numbers of such vehicles on the road increases, it will become more important to manage the charging infrastructure and the electricity distribution network.