The COVID-19 pandemic has severely disrupted travel behavior across diverse socio-economic areas, with a significant impact on transportation systems, public health, and the economy. As countries both recover and plan for future virus-driven stresses, it is crucial to identify the drivers of building travel behavior resilience, such as vaccination. Using an integrated dataset with over 150 million US county-level mobile device data from 01/01/2020 to 20/04/2021, we employ Bayesian structural time series (BSTS) models to infer the relative impact of the vaccination intervention on five types of travel behavior across Metropolitan, Micropolitan and Rural areas. Further, we develop partial least squares regression (PLSR) models to accurately estimate how COVID-19 vaccination rates, epidemiological indicators (i.e., COVID-19 incidence rates, death rates, and testing rates) and weather conditions (i.e., temperature, rain, and snow) would impact various travel behaviors across the diverse areas during the recovery period of the pandemic. The model results shed light on the positive role of vaccinations in fostering the recovery of travel behaviors and reveal the disparities in travel behavior resilience in response to vaccination rates, epidemiological indicators, and weather conditions across diverse areas. Our findings can offer evidential insights for policymakers, transport planners, and public health officials, guiding the development of equitable, sustainable, and resilient transportation systems prepared to adapt to future pandemics.

COVID-19 大流行严重扰乱了不同社会经济领域的出行行为，对交通系统、公共卫生和经济产生了重大影响。随着各国都在复苏并为未来病毒驱动的压力做好规划，确定建立旅行行为弹性的驱动因素（例如疫苗接种）至关重要。使用包含 2020 年 1 月 1 日至 2021 年 4 月 20 日期间超过 1.5 亿美国县级移动设备数据的综合数据集，我们采用贝叶斯结构时间序列 (BSTS) 模型来推断疫苗接种干预对五种类型的相对影响大都市、小都市和农村地区的旅行行为。此外，我们开发了偏最小二乘回归（PLSR）模型来准确估计COVID-19疫苗接种率、流行病学指标（即COVID-19发病率、死亡率和检测率）和天气条件（即气温、降雨、和雪）将影响大流行恢复期间不同地区的各种出行行为。模型结果揭示了疫苗接种在促进旅行行为恢复方面的积极作用，并揭示了不同地区的旅行行为弹性随疫苗接种率、流行病学指标和天气条件的差异。我们的研究结果可以为政策制定者、交通规划者和公共卫生官员提供证据性见解，指导开发公平、可持续和有弹性的交通系统，以适应未来的流行病。