Document Type
Article
Publication Date
6-2024
Keywords
electric vehicles; multi-vehicle household; machine learning; clustering; decision tree; NHTS
Abstract
Electric vehicles (EVs) play a significant role in reducing carbon emissions. In the US, EVs are mostly owned by multi-vehicle households, and their usage is primarily studied in the context of vehicle miles traveled. This study takes a unique approach by analyzing EV usage through the lens of vehicle choice (between EVs and internal combustion engine vehicles) within multi-vehicle households. A two-step machine-learning framework (clustering and decision trees) is proposed. The framework determines the preferred trip category for EV use and captures the effects of household attributes, driver attributes, built-environment factors, and gas prices on EV use in multi-vehicle households. Results indicate that discretionary trips (accumulated local effect = 0.037) are mostly preferred for EV use. EV preference is more pronounced among households with fewer workers (<2) and lower income levels. These findings are valuable for policymakers and auto manufacturers in targeting specific market segments and promoting EV adoption.
Citation
Chowdhury, V., Mitra, S., & Hernandez, S. (2024). Electric Vehicle Usage Patterns in Multi-Vehicle Households in the US: A Machine Learning Study. Sustainability, 16 (12), 5200. https://doi.org/10.3390/su16125200
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.
Included in
Environmental Studies Commons, Science and Technology Studies Commons, Sustainability Commons
