Hogeschool van Amsterdam

Intelligent Data-driven Optimisation of Charging Infrastructure

Predicting charging infrastructure availability based on a space-time series model

Paper

Since the European Union wants to reduce the oil dependence of the transportation system, the uptake of alternative vehicle technologies are stimulated in the member states. In the Netherlands, stimulation is already largely implemented in the form of a comprehensive charging infrastructure. This infrastructure is widely used by the electric vehicle drivers and thus there may occur a form of competition for the charging points. In this paper we address this problem by predicting the short-term availability of charging points at a given location and time by using the historical charging data in a space-time series model. The model shows better accuracy with respect to a naive method for short term predictions up to one day. This will allow charging point operators to provide customers with the service of looking up estimated charging point availability in the nearby future.

Reference De Clerck, Q., van den Hoed, R., & Vanhaverbeke, L. (2016). Predicting charging infrastructure availability based on a space-time series model. Paper presented at 30th International Battery, Hybrid and Fuel Cell Electric Vehicle Symposium, Stuttgart, Germany.
Published by  IDO-Laad 1 January 2016