Ramp metering is one of the most effective tools to combat traffic congestion. In this paper, we present a ramp metering policy for a network of freeways with arbitrary number of on- and off-ramps, merge, and diverge junctions. The proposed policy is designed at the microscopic level and takes into account vehicle following safety constraints. In addition, each on-ramp operates in cycles during which it releases vehicles as long as the number of releases does not exceed its queue size at the start of the cycle. Moreover, each on-ramp dynamically adjusts its release rate based on the traffic condition. To evaluate the performance of the policy, we analyze its throughput, which is characterized by the set of arrival rates for which the queue sizes at all on-ramps remain bounded in expectation. We show that the proposed policy is able to maximize the throughput if the merging speed at all the on-ramps is equal to the free flow speed and the network has no merge junction. We provide simulations to illustrate the performance of our policy and compare it with a well-known policy from the literature.
翻译:在本文中,我们提出了一个用于测量高速公路网络的坡道测量政策,其中任意设置了多处的边际和边际、组合和交汇点。拟议政策是在微观一级设计的,在安全限制之后考虑到车辆。此外,每条边际运行周期期间,只要释放的车辆数量不超过周期开始时的排队规模,就释放车辆。此外,每条边际动态根据交通条件动态调整其释放速度。为了评估政策的执行情况,我们分析了其吞吐量,其特点是所有边际的排队人数仍受预期约束的一套到达率。我们表明,如果所有边际的整合速度与自由流速度相等,而且网络没有连接,则拟议的政策能够最大限度地增加吞吐量。我们提供模拟,以说明我们政策的绩效,并将其与文献中众所周知的政策进行比较。</s>