There is limited availability of travel survey data on households with electric vehicles (EVs) and a lack of evidence on factors influencing EV ownership levels at a finer geographic level, which are crucial for optimizing public charging infrastructure investments. To address this gap, we propose an integrated approach utilizing a discrete choice model and a Bayesian network-generated synthetic population. Applied to Maryland, the model analyzes the impact of public charging stations (level-2 and DC fast chargers) on EV ownership at the census tract level. Access to fast charging, workplace charging, and the possibility of teleworking are key factors influencing EV ownership. The model, applied to the synthetic population, predicts higher EV growth in suburban regions compared to urban areas and a larger increase in EV adoption among high-income groups. This highlights potential disparities in EV adoption and demonstrates the application of this methodology in understanding micro-level EV adoption rates for informing targeted policies and infrastructure development to promote equitable adoption.

中文翻译：

---

充电基础设施对电动汽车采用的影响：综合人口方法


关于拥有电动汽车（EV）的家庭的出行调查数据有限，并且缺乏在更精细的地理层面上影响电动汽车拥有水平的因素的证据，而这对于优化公共充电基础设施投资至关重要。为了解决这一差距，我们提出了一种利用离散选择模型和贝叶斯网络生成的合成群体的综合方法。该模型应用于马里兰州，分析了公共充电站（2 级和直流快速充电器）对人口普查区层面的电动汽车拥有量的影响。快速充电、工作场所充电和远程办公的可能性是影响电动汽车拥有量的关键因素。该模型适用于综合人口，预测郊区地区的电动汽车增长速度高于城市地区，并且高收入群体的电动汽车采用率将大幅增加。这凸显了电动汽车采用方面的潜在差异，并展示了该方法在了解微观层面电动汽车采用率方面的应用，以便为有针对性的政策和基础设施开发提供信息，以促进公平采用。