Dynamic discrete choice modeling of plug-in electric vehicle use and charging using stated preference data.

1438238 (MacKenzie)。缺乏基于经验的充电行为模型日益被认为是现有插电式电动汽车（PEV）研究的局限性。该项目将通过量化多种因素对驾驶员选择的影响来扩展充电行为的知识，这些因素包括：(1)在特定的旅行日使用电动汽车还是替代汽车，以及(2)在一天中的不同时间点是否给电动汽车充电。一个重要的创新在于对这些选择的相互依赖进行建模。数据将使用动态离散选择模型进行分析，该模型在经济学中已经有了相当大的发展，但在运输领域的应用相对较少。研究人员将测试动态选择建模框架是否能比交通工程中常用的静态选择建模方法更好地预测PEV车主对车辆使用和充电的决定。此外，这项工作还将深入了解PEV司机在减少汽油消耗和温室气体排放方面的货币价值，以及这种价值与汽油价格的对比。插电式电动汽车通过减少石油进口支出，使经济免受油价冲击的影响，以及减少国家在地缘政治不稳定地区的战略利益，有助于改善国家安全和提高经济竞争力。它们通过减少交通对当地和全球环境质量的影响，提高社会中个人的福祉。这项研究将通过(1)支持更好地匹配消费者实际行为的车辆和充电基础设施网络的设计，以及(2)允许政策制定者基于对其潜在效益的更准确评估，为pev和充电基础设施设计具有成本效益的激励措施，从而促进向更具环境和经济可持续性的交通系统的转变。这项研究将通过强调现代工程师多学科能力的重要性来丰富STEM教育。PI计划将这项研究的成果纳入他正在开设的一门关于交通能源和可持续发展的新课程的作业和材料中。研究人员拥有跨越公共、私人和学术部门的专业网络，并致力于将通过本研究创造的知识传授给实践者。他们预计这项工作将为汽车制造商和充电网络运营商带来好处，更有效的政策制定，并加强这些群体之间的关系。一个声明的偏好选择实验将通过一个定制的、基于网络的界面对PEV司机样本进行管理。该调查将收集人口统计和车辆拥有量方面的关键背景信息，然后向受访者提供一套量身定制的选择方案。受访者将获得一个计划的旅行日，其中包括一个或多个驾驶段（指定距离），其中穿插着潜在的充电机会（以停车时间长短、成本、电力和充电设备的典型可用性为特征）。他们将被问及在指定的旅行日是否使用自己的电动汽车或替代车辆。那些表示他们会使用电动汽车的人将被带到一个模拟旅行日的过程中，在这个过程中，他们将被问及是否会在每次机会时给他们的汽车充电。调查数据将使用动态离散选择建模方法进行分析。在这个框架中，在较早时期做出的选择会影响随后选择的潜在回报，并且假设较早的选择是在这种预见的情况下做出的。PI假设这种建模方法可以更好地代表PEV驾驶员的决策过程，在决定是否在当天早些时候驾驶或充电时，他们会考虑在当天晚些时候找到适当充电机会的可能性。这项工作的一个关键成果将是一个PEV驾驶员选择使用和充电PEV的概率模型，该模型的条件是（可能不确定的）有关出行计划和后续充电机会的信息。

1438238 (MacKenzie). The lack of empirically based models of charging behavior is increasingly acknowledged as a limitation of existing Plug-in Electric Vehicle (PEV) research. This project will expand knowledge of charging behavior by quantifying the effects of multiple factors on drivers' choices over (1) whether to use a PEV or an alternate vehicle for a particular travel day, and (2) whether or not to charge their PEV at various points throughout their day. An important innovation lies in modeling the interdependence of these choices. Data will be analyzed using dynamic discrete choice models, which have seen considerable development in economics but relatively little use in transportation. The investigators will test whether the dynamic choice modeling framework can better predict PEV owners' decisions about vehicle use and charging than can the static choice modeling approaches more commonly used in transportation engineering. Additionally, the work will provide insights into the monetary value that PEV drivers place on reducing gasoline consumption and greenhouse gas emissions, and how this value compares to the price of gasoline. Plug-in electric vehicles contribute to improved national security and enhanced economic competitiveness by reducing spending on oil imports, insulating the economy from oil price shocks, and reducing the country's strategic interests in geopolitically unstable regions. They enhance the wellbeing of individuals in society by reducing the impacts of transportation on local and global environmental quality. This research will facilitate a shift to a more environmentally and economically sustainable transportation system by (1) supporting the design of vehicles and charging infrastructure networks that better match consumers' actual behavior, and (2) allowing policymakers to design cost-effective incentives for PEVs and charging infrastructure, based on more accurate assessments of their potential benefits. This research will enrich STEM education by highlighting the importance of multidisciplinary competence for the modern engineer. The PI plans to incorporate products from this research into assignments and materials for a new course he is offering on Transportation Energy & Sustainability. The investigators have professional networks spanning the public, private, and academic sectors, and are committed to transferring to practitioners the knowledge created through this research. They anticipate benefits for automobile manufacturers and charging network operators, more effective policymaking, and strengthened relationships among these groups to result from this work. A stated preference choice experiment will be administered via a customized, web-based interface to a sample of PEV drivers. The survey will collect key background information on demographics and vehicle ownership, then present respondents with a set of tailored choice situations. Respondents will be presented with a planned travel day that includes one or more segments of driving (of specified distances) interspersed with potential charging opportunities (characterized by the length of the stop and the cost, power, and typical availability of charging equipment). They will be asked whether they would use their PEV or an alternate vehicle for the specified travel day. Those who indicate that they would use their PEV will be taken through a simulation of the travel day, in which they will be asked whether or not they would charge their vehicle at each opportunity. The survey data will be analyzed using a dynamic discrete choice modeling approach. In this framework, choices made in earlier periods affect the potential payoffs of subsequent choices, and the earlier choices are assumed to be made with this foresight in mind. The PI hypothesizes that this modeling approach can better represent the decision processes of PEV drivers, who consider the likelihood of finding appropriate charging opportunities later in the day when deciding whether to drive or charge a PEV earlier in the day. A key product of this work will be a model of the probability of a PEV driver choosing to use and charge a PEV, conditional on (potentially uncertain) information about travel plans and subsequent charging opportunities.