A Stochastic Model for Uncontrolled Charging of Electric Vehicles Using Cluster Analysis

Constance Crozier,Thomas Morstyn,Malcolm McCulloch

DOI: https://doi.org/10.48550/arXiv.1907.09458

2019-07-17

Abstract:This paper proposes a probabilistic model for uncontrolled charging of electric vehicles (EVs). EV charging will add significant load to power systems in the coming years and, due to the convenience of charging at home, this is likely to occur in residential distribution systems. Estimating the size and shape of the load will allow necessary reinforcements to be identified. Models predicting EV charging are usually based on data from travel surveys, or from small trials. Travel surveys are recorded by hand and typically describe conventional vehicles, but represent a much larger and more diverse sample of the population. The model here utilizes both sources: trial data to parameterize the model, and survey data as the model input. Clustering is used to identify modes of vehicle use, thus reducing vehicle use to a single parameter -- which can be incorporated into the model without adding significant computational burden. Two case studies are included: one investigating the aggregated charging of 50 vehicles, and one predicting the increase in after diversity maximum demand for different regions of the UK.

Signal Processing