Intelligent Transportation Planning

如今，公路运输部门是温室气体排放的一个主要因素。这种贡献主要是由于传统化石燃料驱动的车辆占很大比例，以及道路上车辆数量的不断增加。然而，交通运输部门正在经历一场真正的技术革命，被称为交通运输4.0 （Tr4)，其主要要素是：1)向交通运输系统电气化的过渡，2)采用自动驾驶和自动驾驶汽车（AVs）， 3）信息技术的进步，4)交通数据的广泛收集和存储，以及5)对计算资源的即时和分布式访问。虽然Tr4中隐含的技术进步为更可持续和高效的交通系统创造了条件，但它们的管理变得更加具有挑战性。在调查科学文献时，人们发现，无论是在运输系统的背景下，还是在货运分配方面，都相当缺乏解决Tr4带来的挑战的方法。因此，本建议的主要目标是开发决策支持工具，使运输行业能够充分利用Tr4提供的机会。所提出的方法根植于运筹学和人工智能领域。本研究计划围绕两个主题展开。第一个主题考虑到运输系统，并着重于按需服务与传统运输结合所产生的规划问题。这是通过采用名为半灵活运输系统（STS）的建模框架来实现的。运输服务系统的目标是在定期运输的框架内提供响应需求的服务。综合了传统系统和按需系统的特点，技术支助系统需要极其复杂的规划活动。更具体地说，本提案旨在开发将自动驾驶车队纳入公交系统运营规划的方法，并构建仿真框架来分析公交系统与传统交通系统集成的性能及其对城市尺度交通的影响。第二个主题考虑货运，并着重于Tr4背景下出现的随机和动态路由问题。更具体地说，第二个主题旨在开发在随机和动态交通条件下车辆实时路径的决策支持工具，以及在运行配备电池的电动汽车车队时的路径和充电操作的动态政策。通过开发工具来实现Tr4革命，本研究项目的预期影响是提供更可持续和更公平的交通服务，无论是在公共交通方面，还是在货运分配方面。

The road-transportation sector represents nowadays a major contributing factor to the count of Green-House Gas emissions. This contribution is mostly due to the large ratio of vehicles powered by traditional fossil fuels, and by the constant rise of the number of vehicles on the roads. The transportation sector has, however, the perspective of undergoing an authentic technological revolution, labelled as Transportation 4.0 (Tr4), whose main elements are: 1) the transition towards the electrification of the transportation system, 2) the adoption of self-driving and Autonomous Vehicles (AVs), 3) the advances in Information Technology, 4) the widespread collection and storage of transportation data, and 5) the instantaneous and distributed access to computing resources. While the technological advances implicit in the Tr4 set up the conditions for more sustainable and efficient transportation systems, their management becomes dramatically more challenging. When surveying the scientific literature, one finds that there is a considerable lack of methodologies to address the challenges presented by the Tr4, both in the context of transit systems, as well as in freight distribution. Therefore, the main objective of this proposal is to develop decision-support tools to enable the transportation industry to fully profit from the opportunities given by the Tr4. The proposed methods are rooted in the fields of operations research and artificial intelligence. This research program is organized around two themes. The first theme considers transit systems and focuses on planning problems arising in the integration of on-demand services with traditional transit. This is done by adopting a modeling framework named Semi-flexible Transit System (STS). STSs aim to provide demand-responsive service within the framework of regularly scheduled transit. Combining characteristics of traditional and on-demand systems, STSs require an extremely complex planning activity. More specifically, this proposal aims to develop the methodology to incorporate a fleet of AVs in the operations planning of STSs, and build a simulation framework to analyze the performance of a system integrating STSs with traditional transit, and its impact on the mobility at a city-scale. The second theme considers freight transportation and focuses on stochastic and dynamic routing problems arising in the context of the Tr4. More specifically, this second theme aims to develop decision-support tools for real-time routing of vehicles under stochastic and dynamic traffic conditions, as well as dynamic policies for routing and charging operations when operating a fleet of battery-equipped electric vehicles. By developing tools to enable the Tr4 revolution, the anticipated impact of this research program is to provide more sustainable and more equitable transportation services, both in terms of public transit, as well as freight distribution.