Analyzing locational inequalities in the placement of electric vehicle charging stations using machine learning: A case study in Glasgow
Year: 2024
Type: article
Abstract: This study introduces a comprehensive machine learning framework aimed at improving the spatial accessibility of Electric Vehicle Charging Stations (EVCS) in Glasgow by identifying optimal locations for new placements. Addressing the growing need for equitable Electric Vehicle (EV) infrastructure, the framework integrates a diverse range of socioeconomic, demographic, and environmental data with historical EV charging patterns to accurately predict areas requiring infrastructure investment. Three machine learning models—Support Vector Machines (SVM), Multinomial Logistic Regression (MLR), and Random Forest (RF)—were trained and tested across various spatial scales to assess their effectiveness in capturing the spatial and socioeconomic variability of EVCS needs across Glasgow. Among these models, Random Forest achieved the highest cross-validation accuracy of 95.6% at a 3 km grid resolution, making it the most effective model for predicting future EVCS placements. Unlike prior research, which often focuses on cost optimization or lacks consideration of social equity, this study develops the Electric Vehicle Charging Inequity Index (EVCI), a novel tool combining indicators like income, population density, and minority status to identify high-inequity zones. The EVCI highlighted specific areas in central Glasgow, including Pollokshields, Govanhill, and the city center, as high-inequity zones with limited access to EV charging stations. Our analysis showed that approximately 11.04% of the city falls into high-inequity zones, where EVCS access is minimal. Medium-inequity areas, comprising 69.52% of the city, have moderate access to charging infrastructure, while low-inequity zones, representing 19.11% of Glasgow, enjoy relatively good EVCS availability. This framework provides urban planners and policymakers with actionable insights into where EVCS resources should be allocated to support equitable access to EV charging across Glasgow. By identifying high-inequity areas in need of infrastructure, this model promotes the efficient and targeted expansion of EVCS, closing accessibility gaps and fostering an inclusive, sustainable transition to electric vehicles in urban areas. The adaptable nature of the framework allows it to be applied to other cities, offering a replicable approach for assessing and addressing spatial inequities in EV infrastructure deployment.
Source: Next research.
Author: Arash Mousaei
Institution: University of Tabriz
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $2,500
OpenAlex ID: W4405910854