Assessing monthly utilization of public electric vehicle charging infrastructure using weighted averaging strategies for a low-carbon transportation system

Abstract

The rapid growth of electric vehicles (EVs) holds promise for achieving carbon neutrality, and public charging infrastructure is a vital component of low-carbon transportation solutions. Existing assessment frameworks for public EV charging infrastructure utilization primarily emphasize hourly or daily analyses along with a simple averaging method. However, there has been little attention paid to monthly analysis and weighted averaging methods. To address this gap, this study introduces three distinct monthly utilization measures and three innovative weighted averaging strategies, leveraging 13.7 million data sessions over 4 years. Two preprocessing steps, including outlier detection based on meta-analysis and identification of incomplete initial month data, ensure reliable monthly utilization from large-scale data. The study then suggests three monthly utilization metrics to holistically assess public charger operation through the number of charging sessions, charging demand, and charging time, rather than relying on a single metric. Based on these metrics, the trends and distributions of monthly utilization are explored across charging options. A downtime estimation algorithm is developed to exclude non-operational periods from average utilization calculations for EV chargers. Differing from the simple averaging method commonly applied in current works, this study introduces three weighting schemes, namely, Gaussian, temporal, and combination weighting. The results show that the Gaussian weighting offers robustness to abrupt variability, the temporal weighting places greater emphasis on recent trends, and the combination strategy merges the strengths of the Gaussian weighting and the temporal weighting.