CAREER: Impact of Freeway Geometric Design on Congestion Characteristics

该学院早期职业发展（Career）项目奖的目标是建立高速公路设计元素（如坡度和曲率）与交通拥堵特征（如走走停停的频率、迟滞和容量下降）之间的关系。研究方法是基于实证数据分析，既有现有的和收集作为这个项目的一部分。车辆轨迹数据将用于通过数值模拟和统计分析相结合来关联拥堵特征和个体驾驶员行为。这将揭示驾驶员行为与高速公路设计之间的联系，并提出并验证能够捕捉本研究结果的汽车跟随理论。如果成功，这项研究的结果将导致改进高速公路设计指南，目前主要侧重于提供足够的容量。这项研究将为以下问题提供答案：我们应该如何设计高速公路，以便在拥堵时，(i)加速/减速机动的数量最小化？或（ii）瓶颈排放率最大化？，或（iii）尽量减少变道次数？这些问题的答案可能有助于减少温室气体排放、通勤延误和事故，并将在未来促进一个更可持续的基础设施系统。此外，本研究将为从业人员和研究人员提供第一个能够捕捉经验观察到的复杂拥堵现象的交通流模型。这将提高现有高速公路基础设施项目模拟评估的准确性，并将允许测试更有效的拥堵控制策略，重点是尽量减少排放水平、延误和事故。教育和推广目标包括一个促进交通流理论、模拟和建模的在线门户网站，学生、研究人员和行业专业人士将从中受益。该门户网站包括每周网络研讨会、数据和程序工具、讨论论坛和wiki。

The objective of this Faculty Early Career Development (CAREER) program award is to establish the relationship between freeway design elements, such as grade and curvature, and traffic congestion characteristics, such as the frequency of stop-and-go movements, hysteresis and capacity drop. The research approach is based on empirical data analysis, both existing and collected as part of this project. Vehicle trajectory data will be used to correlate congestion characteristics and individual driver behavior using a combination of numerical simulations and statistical analyses. This will allow for the unveiling of the link between driver behavior and freeway design, and to propose and validate a car-following theory able to capture the findings in this research. If successful, the results of this research will lead to improved freeway design guidelines, which currently mostly focus on providing adequate capacity. This research will provide answers to questions such as: How should we design freeways so that, when congested, (i) the number of acceleration/deceleration maneuvers is minimized?, or (ii) the bottleneck discharge rate is maximized?, or (iii) the number of lane-changes is minimized? The answer to these questions will potentially help minimize greenhouse gases, commuter delays and accidents, and will promote a more sustainable infrastructure system in the future. Additionally, this research will provide practitioners and researchers with arguably the first traffic flow model able to capture complex congestion phenomena as observed empirically. This will improve the accuracy of existing freeway infrastructure project simulation-based assessments, and will allow testing more efficient congestion control strategies that focus on minimizing emission levels, delays and accidents. The educational and outreach objectives include an on-line portal for promoting traffic flow theory, simulation and modeling, where a community of students, researchers and industry professionals will benefit. This portal includes weekly webinars, data and program tools, a discussion forum and a wiki.