文章编号：1672-5328（2012）06-0090-05

Pravin Varaiya¹　著，吕大玮²　译
（1.加利福尼亚大学电气工程与计算机科学系，加利福尼亚州 伯克利 94720；2.中国城市规划设计研究院，北京 100037）

摘要：随着高速公路监测手段的完善，高速公路运行数据日益丰富。通过深入挖掘、分析加州高速公路运行监测系统5年来收集的数据，从不同角度定量理解拥堵产生的原因，包括拥堵动态分析、交通瓶颈识别、匝道控制效益评价、出行时间预测等。此外，基于上述数据，以加州湾区HOV车道为例，通过分析通行能力损失、对比路段相同位置HOV限制与非限制时段车道流量及速度变化，定量评价了HOV车道的实施效果及其给其他车道带来的拥堵后果。上述各项研究均衡量了拥堵的严重程度及产生原因，并给出了相应的缓堵方法。

关键词：高速公路；交通拥堵；HOV车道；匝道控制；实时数据

中图分类号：U491

文献标识码：A

What We’ve Learned about Highway Traffic Congestion

Written by Pravin Varaiya¹, Translated by LÜ Dawei2
(1.Department of Electrical Engineering and Computer Science, University of California, Berkeley CA 94720, USA; 2. China Academy of Urban Planning & Design, Beijing 100037, China)

Abstract：Freeway operational data become more available and enriched with the improvement of surveillance systems. By analyzing five years of freeway operational data archived in California, this paper quantitatively elaborates the causes of highway congestion from different perspectives, including congestion dynamics analysis, traffic bottleneck identification, ramp metering performance evaluation, travel time prediction, and etc. In addition, this paper quantifies the performance of HOV lanes and their impacts on other lanes in California Bay Area by analyzing capacity loss and comparing the speed and flow within and outside of HOV restriction periods for the same highway segment. Each of the aforementioned studies measures the severity and cause of congestion, and suggests mitigation approaches.

Keywords：freeway; highway congestion; HOV lane; ramp metering; real-time data