Collaborative Research: Mixed Traffic Dynamics Under Disturbances: Impact of Multi-Class Connected and Automated Vehicles

合作研究：干扰下的混合交通动态：多类互联和自动驾驶车辆的影响

• 批准号: 1932921
• 负责人: Danjue Chen
• 金额: $ 9.99万
• 依托单位: University of Massachusetts Lowell
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-11-01 至 2022-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1932921&HistoricalAwards=false
• 关键词: Collaborative; Research; Mixed; Traffic; Dynamics

Connected and Automated Vehicle (CAV) technologies have garnered huge interest across private industry, academia, government, and the public. A wide range of benefits are predicted when these ground-breaking technologies become mature, including higher road efficiency, improved safety, and better energy consumption and emissions. However, these benefits will be open to question until the technologies sufficiently mature. Specifically, a major uncertainty in benefits lies in mixed traffic of CAVs and human-driven vehicles (HDVs), where interactions between them remain largely unknown. Therefore, in the foreseeable future, traffic will likely be mixed with multiple classes of CAVs and HDVs. This project will aim to better understand the anticipated behavior of this mixed traffic system, and its impact on traffic in order to help fully utilize the potentials of the CAV technology. The results will guide the development of traffic management strategies, policies, and long-term planning for the future transportation system. This project will also engage in a range of integrated research, educational and outreach activities that will extend the knowledge obtained from this research to a broader audience, including developing simulation-based educational modules, organizing workshops, sharing simulation platform for mixed traffic, and engaging undergraduate and graduate students, particularly underrepresented groups, in the research and education.This research aims to understand how HDVs and different classes of CAVs will interact under traffic disturbances that cause (momentary) reductions in speed and affect traffic flow performance. Specifically, this project will aim to (1) characterize discernable differences in the car-following behavior of HDVs and CAVs of different classes under disturbances; and (2) elucidate their effects on traffic flow throughput and traffic flow stability. To this end, this research will develop a systematic method to bring together different control modeling paradigms for CAVs into a unifying framework to unveil their individual and collective impacts on traffic flow throughput and stability. Three CAV control paradigms will be considered in this study: linear control, model predictive control (MPC), and artificial-intelligence-based control. The vehicle-level investigation of complex interactions among CAVs and HDVs will unveil the interaction mechanisms and elucidate how they scale up to the collective behavior of traffic stream, which will inspire new modeling paradigms to describe mixed traffic flow dynamics and control CAVs.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

互联和自动驾驶汽车（CAV）技术已经引起了私营行业、学术界、政府和公众的巨大兴趣。当这些突破性的技术成熟时，预计会带来广泛的好处，包括更高的道路效率，更好的安全性以及更好的能源消耗和排放。然而，在技术足够成熟之前，这些好处将是值得商榷的。具体而言，效益的主要不确定性在于CAV和人类驾驶车辆（HDV）的混合交通，其中它们之间的相互作用在很大程度上仍然未知。因此，在可预见的未来，交通可能会与多种CAV和HDV混合。该项目旨在更好地了解这种混合交通系统的预期行为及其对交通的影响，以帮助充分利用CAV技术的潜力。研究结果将指导未来交通系统的交通管理战略、政策和长期规划的制定。该项目还将开展一系列综合研究、教育和外联活动，将从这项研究中获得的知识扩展到更广泛的受众，包括开发基于模拟的教育模块，组织讲习班，共享混合交通模拟平台，并吸引本科生和研究生，特别是代表性不足的群体，本研究旨在了解HDV和不同类别的CAV在导致（瞬时）速度降低并影响交通流性能的交通干扰下将如何相互作用。具体而言，本项目的目标是（1）表征在扰动下不同类别的HDV和CAV的跟车行为的可辨别差异;（2）阐明它们对交通流吞吐量和交通流稳定性的影响。为此，本研究将开发一种系统的方法，将不同的控制建模范式的CAV到一个统一的框架，揭示他们的个人和集体对交通流量的吞吐量和稳定性的影响。三个CAV控制范例将被认为是在这项研究中：线性控制，模型预测控制（MPC），和人工智能为基础的控制。对CAV和HDV之间复杂相互作用的车辆级研究将揭示相互作用机制，并阐明它们如何扩展到交通流的集体行为，该奖项反映了美国国家科学基金会的法定使命，并被认为是值得通过利用基金会的智力价值和更广泛的影响评审进行评估来支持的的搜索.

项目成果

Dampen the Stop-and-Go Traffic with Connected and Automated Vehicles – A Deep Reinforcement Learning Approach*

• DOI: 10.1109/mt-its49943.2021.9529289
• 发表时间: 2020-05
• 期刊: 2021 7th International Conference on Models and Technologies for Intelligent Transportation Systems (MT-ITS)
• 作者: Liming Jiang;Yuanchang Xie;Danjue Chen;Tienan Li;Nicholas G. Evans

A deep reinforcement learning based distributed control strategy for connected automated vehicles in mixed traffic platoon

• DOI: 10.1016/j.trc.2023.104019
• 发表时间: 2023-03
• 期刊: Transportation Research Part C: Emerging Technologies
• 作者: Haotian Shi;Danjue Chen;Nan Zheng;Xin Wang;Yang Zhou;Bin Ran

Processing, assessing, and enhancing the Waymo autonomous vehicle open dataset for driving behavior research

• DOI: 10.1016/j.trc.2021.103490
• 发表时间: 2022-01-01
• 期刊: TRANSPORTATION RESEARCH PART C-EMERGING TECHNOLOGIES
• 影响因子: 8.3
• 作者: Hu, Xiangwang;Zheng, Zuduo;Sun, Jian
• 通讯作者: Sun, Jian

On multi-class automated vehicles: Car-following behavior and its implications for traffic dynamics

• DOI: 10.1016/j.trc.2021.103166
• 发表时间: 2021-07
• 期刊: Transportation Research Part C: Emerging Technologies
• 作者: Wissam Kontar;Tienan Li;A. Srivastava;Yang Zhou;Danjue Chen;Soyoung Ahn

Fundamental Diagrams of Commercial Adaptive Cruise Control: Worldwide Experimental Evidence

• DOI: 10.1016/j.trc.2021.103458
• 发表时间: 2021-05
• 期刊: ArXiv
• 作者: Tienan Li;Danjue Chen;Hao Zhou;Jorge A. Laval;Yuanchang Xie