Adaptive Traffic Signal Control: A Novel Modelling Approach

Abstract

Deploying charging infrastructure requires identifying the most effective placement and size of charging facilities. This is particularly challenging when electric vehicles (EVs) are gradually introduced, as it creates a dynamic target that must be met to ensure successful adoption of EVs. This paper introduces an agent-based simulation model that tracks movements of EVs in space and time. Our model is based on a choice model of charging behaviour, which is integrated with non-parametric queues and information-sharing of waiting time. Our simulation captures demand resulting from choice of charging and the unserved demand represented as charging intentions that were not met. It is demonstrated that unserved demand varies over time and across locations, and that it can be greatly reduced by our information-sharing strategy. The model is applied to Copenhagen where we examine changes in charging infrastructure requirements between 2021-2030 when going from EV shares of 2% to 30%.