Abstract

Ramp metering is a major measure to reduce traffic congestion and prevent traffic breakdown at on-ramp bottlenecks. However, its efficiency is often limited by the excessive traffic demands on either the mainline or the on-ramp. Meanwhile, unequal lane-setting, in which the number of mainline lanes is more than on-ramp, is a common infrastructure setup in urban expressway system. In this paper, a novel Travel time balanced Mainline and Ramp integrated Control strategy (T-MRC) is proposed for unequal lane-setting on-ramp bottlenecks, which follows the feedback control framework. The control goal is to maximize the traffic efficiency, while in view of the typical early-onset phenomenon, productivity (efficiency index) is selected to replace the conventional outflow as the optimization objective. On this basis, the balance of travel time in different types of lanes (controlled on-ramp, controlled, and uncontrolled mainline lanes) is considered, and two regulators are used in the green time allocation algorithm. A high-precision microscopic simulation model in Shanghai is set up, and five control strategy scenarios are compared. Results indicate that the T-MRC strategy outperforms all others that the productivity is improved dramatically (>85%) and the vehicle total delay (VTD) of the system is reduced significantly (>30%) compared to no-control scenario.

摘要

匝道计量是减少交通拥堵和防止匝道瓶颈处交通中断的主要措施。然而，它的效率通常受到主线或入口匝道上过多的流量需求的限制。同时，主线车道数多于匝道的不等车道设置是城市高速公路系统中常见的基础设施设置。在本文中，针对不等车道设置的匝道瓶颈提出了一种新颖的行驶时间平衡主线和匝道集成控制策略（T-MRC），该策略遵循反馈控制框架。控制目标是最大化交通效率，同时针对典型的早发现象，选择生产力（效率指数）代替常规流出作为优化目标。以这个为基础，考虑了不同类型车道（受控入口匝道、受控和非受控主线车道）的行驶时间平衡，在绿灯时间分配算法中使用了两个调节器。在上海建立了高精度的微观仿真模型，并比较了五种控制策略场景。结果表明，与无控制场景相比，T-MRC 策略优于所有其他策略，生产率显着提高（> 85%），系统的车辆总延迟（VTD）显着减少（> 30%）。