A User-Friendly Electric Vehicle Reallocation Solution for Uniformly Utilized Charging Stations

Abstract

Electric vehicle (EV) users often struggle to identify charging stations (EVCSs) that are both available at the time of interest and placed in convenient locations. At the same time, EVCSs in nearby locations may systematically be underutilized. This variance in the utilization of the EVCSs can have negative implications, both for the EV users (e.g. higher anxiety, waiting time, and discomfort) and the owners/operators of the EVCSs (e.g. lower income and longer payback period of investment). The present work aims to mitigate said implications by introducing a solution that efficiently allocates EV users to stations based on their expected utilization and their proximity to the charging locations of preference. To do so, a forecasting model is first used to predict the arrivals at the EVCSs. Then, the utilization rates of the EVCSs are computed and a heuristic algorithm is employed to dynamically reallocate the EV users from overutilized stations to nearby, underutilized stations. The proposed approach, which has been implemented in a web-based application, is evaluated through simulation experiments and compared with the traditional ``first-come, first-served'' approach. The results indicate that our approach can achieve more uniformly distributed utilization rates with a minimal deterioration on the convenience of the EV users, thus offering a practical solution to the examined problem.

Keywords: Electric vehicles, Charging station utilization, Optimization, forecasting, Web application