Socially optimal charging strategies for electric vehicles

TL;DR

This paper develops decentralized charging policies for electric vehicles to minimize queues, analyzes their social optimality in multi-server settings, and offers insights into infrastructure deployment for improved EV adoption.

Contribution

It introduces decentralized policies that ensure social optimality for EV charging, with convergence analysis and infrastructure deployment insights.

Findings

Policies achieve minimal queueing in multi-server regimes

Convergence of the optimal policy algorithm is established

Guidelines for optimal charging station deployment are provided

Abstract

Electric vehicles represent a promising technology for reducing emissions and dependence on fossil fuels and have started entering different automotive markets. In order to bolster their adoption by consumers and hence enhance their penetration rate, a charging station infrastructure needs to be deployed. This paper studies decentralized policies that assign electric vehicles to a network of charging stations with the goal to achieve little to no queueing. This objective is especially important for electric vehicles, whose charging times are fairly long. The social optimality of the proposed policies is established in the many-server regime, where each station is equipped with multiple charging slots. Further, convergence issues of the algorithm that achieves the optimal policy are examined. Finally, the results provide insight on how to address questions related to the optimal location deployment of the infrastructure.