Household-Level Use of Autonomous Vehicles: Modeling Framework, Traveler Adaptation, and Infrastructure to Mitigate Negative Effects

随着车辆自动化的发展，与自动驾驶车辆（AV）或自动驾驶车辆共享公共道路的想法已被广泛接受。自动驾驶汽车的采用将彻底改变交通系统的现状。预计将发生巨大变化的领域之一是车辆使用，出行行为，以及随后的无人驾驶操作的车辆所有权。然而，目前还没有一个框架来预测这种行为、需求和系统性变化。首先，这项研究将定义一个框架，以模拟和评估潜在的家庭层面的使用自动驾驶汽车的基本概念，旅行是一个派生的需求，个人的活动和时间分配。第二，本研究将制定一个基础设施规划框架，重点是停车结构规划，以减少潜在的负面外部影响，如增加车辆行驶里程（VMT），拥堵，温室气体排放等。本研究将提高对潜在问题和外部影响的理解，以及政策和基础设施规划策略，以帮助降低自动驾驶汽车采用的社会成本。更广泛地说，这项研究将有助于通过自动驾驶汽车的采用创造一个更可持续、更经济、更高效的交通和基础设施系统。自动驾驶操作的框架是车辆路径问题（VRP）的创新变体，因为（1）它集成了旅行者和车辆的两个调度和路径问题，以及（2）结合了无人驾驶操作。这种提法的新奇在于，它集成了两个路由问题，从而为旅行者和车辆提供路由的灵活性。基础设施规划框架通过双层编程建模，其中AV操作决策在每个家庭级别。一个创造性的建模方法来处理计算的挑战。最后，敏感性分析将有助于了解采用自动驾驶汽车可能带来的基础设施、环境、拥堵、车辆所有权等方面的变化和挑战。

With recent development in vehicle automation, the idea of sharing public roads with Autonomous Vehicles (AVs) or self-driving vehicles has been widely accepted. Adoption of AVs will drastically change the current picture of the transportation system. One of the areas that is expected to be greatly transformed is vehicle use, travel behavior, and subsequently vehicle ownership from the driverless operations. Yet, there exists no framework to forecast such behavior, demand, and systemic changes at this stage. First, this research will define a framework to simulate and evaluate potential household-level use of AVs from the fundamental notion that travel is a derived demand of individuals' activities and time allocations. Second, this research will develop an infrastructure planning framework, with a focus on parking structure planning, to reduce the potential negative external effects such as increased Vehicle-Mile-Traveled (VMT), congestion, greenhouse gas emissions, etc. The research will enhance understanding of potential issues and external effects as well as policy and infrastructure planning strategies to help reduce the social cost of AV adoption. More broadly, this research will help creating a more sustainable, economic, and efficient transportation and infrastructure system through the AV adoption. The framework of AV operations is an innovative variation of the Vehicle Routing Problem (VRP) in that (1) it integrates two scheduling and routing problems of travelers and vehicles, and (2) incorporates driverless operations. The novelty of this formulation is that it integrates two routing problems, thus providing routing flexibility for travelers and vehicles. The infrastructure planning framework is modeled through bi-level programming where AV operation decisions are at each household level. A creative modeling approach is introduced to handle the computational challenges. Lastly, a sensitivity analysis will provide insights in understanding changes and challenges in infrastructure, environment, congestion, vehicle ownership, etc. that may arise with the adoption of AVs.