基于交通流理论的公交站点间行程时间预测

分析了公交站点间车辆运行过程,将行程预测时间划分为交叉口排队等待时间、路段行驶时间和停站时间3个部分,利用交通波理论和延误三角形,分别建立了无公交专用车道和有公交专用车道2种情况下排队等待时间的动态预测模型;根据乘客到站规律和上下车规律,提出了公交车进站停靠时间模型;针对无公交专用车道条件下的时间预测方法进行了实例演算.实验数据表明,基于交通波行程时间预测方法具有较高的精度,可以满足站点间行程时间预报要求.     

快速路菱形立交出入口间距设置及适应性研究

应用Vissim仿真模型研究了在不同的主辅路流量组合和交织流量比下,菱形立交出入口区间平均速度、平均最大排队长度、出口流量与出入口间距的关系.研究表明,交织区平均速度、平均最大排队长度和出口流量都随出入口间距的变化而变化.当主辅路流量和交织流量比较小时,三者随出入口间距变化较小;当主辅路流量和交织流量比较大时,三者随出入口间距变化较大.因此可以得出,当交通流量较小或出入口间距受限时,出入口间距为l00 m时比较适宜;当交通流量较大时,随交织流量比由0.2增大到0.3以上时,出入口间距应由300 m增大到450m,才能有较好的运行状态.     

Adaptive road border crossing system management: the role of priority programs

Land border crossings in North America, such as those between Canada and U.S.A., are expected to experience severe imbalance of travel demand and capacity of processors. During peak travel periods, this is already the case at high traffic locations. The land border crossing authorities have to address problems of congestion, national security and environmental impacts in the operation of the existing systems and to continue to address these problems as a part of infrastructure expansion plans. There is a need to adapt the crossing system management in order to accommodate efficiency and productivity‐oriented priority crossing measures. From a methodological perspective, it is a challenge to evaluate the role of priority crossing measures within the complex border crossing system. This paper reports research on modelling priority crossing initiatives. A microsimulation approach was used to model and analyse integrated processors of the Peace Bridge crossing system between Fort Erie (Ontario) and Buffalo (New York) under different scenarios of travel demand, customs processing times, priority crossing and queue jump lanes for automobile and truck traffic. Findings show the extent to which a border crossing system with priority crossing and queue jump lanes is more efficient and productive than one without these innovations. Copyright © 2011 John Wiley & Sons, Ltd.     

基于视频图像处理的车辆排队长度检测

综合灰度检测和边缘检测的优点，提出了一种将灰度与边缘检测相结合的方法来检测队列长度，采用帧平均法来处理视频流，可以减小由于摄像头抖动或背景微小变化而产生的误差。提出的队列检测算法包含运动检测和车辆存在检测，通过合理分配使用这两种操作，节省了运算时间。实验结果表明，此方法实现简单，能够获得较满意的检测结果。     

信号交叉口排队长度预测的神经网络方法

预测信号交叉口的排队长度可以为交通信号控制和管理提供非常重要的信息。基于神经网络，针对定时和感应信号交叉口两种不同情况，成功实现了单、双车道排队长度的预测。同时，感应器与停车线之间的距离对预测精度的影响也进行了初步分析。模拟结果同时表明，神经网络对左转车道排队长度的预测效果不佳，不能为信号控制和管理提供有效的信息。     

排队论在成渝高速公路收费站服务台设计和管理中的应用

运用排队论对高速公路收费站服务强度、队列中排队等待的平均车辆数及车辆在收费站的平均逗留时间、车辆平均排队等待时间进行分析,建立成渝高速公路收费站服务台与工作人员配备模型,并用实例证明此模型的适用性,对于减少运营成本,提高服务水平效果显著.     

考虑拖延行为的排队长度模型研究

HCM2000 Back of Queue模型(BQ模型)的计算值比实际调查值偏大,而且模型关于排队车辆的界定不能保证现场调查的客观性。提出了车辆拖延行为的定义,分析了拖延行为使BQ模型偏大的2种情况。建立了拖延行为模型,将车辆减速-停车等待-加速过程转化为温和减速-低匀速-温和加速的3阶段模型。通过低匀速行驶时间与停车时间之间的关系,可以得到车辆发生拖延行为的条件。建立了考虑拖延行为的排队长度与BQ模型之间的关系,并以首辆拖延车辆能否通过交叉口以及饱和度是否大于1为判别条件,提出了排队长度修正模型。通过实例验证了修正模型的可行性,而且修正模型可以保证现场调查的客观性和可操作性。     

考虑排队阻滞和换道的改进宏观交通流模型

考虑路段上车辆换道及交叉口处排队阻滞作用,对交通流宏观演化模型进行改进.首先在排队中考虑不同车道的速度、密度差异,建立了换道模型;然后在交叉口内部划分各流向排队车辆的累积路径,分析不同的排队溢出位置对其他车流的阻滞影响;综合考虑车辆换道及交叉口处阻滞影响,建立了改进的车道组宏观交通流模型;最后设计仿真实验对模型进行验证.结果表明,在高流量需求下,本文模型能够模拟出下游的排队溢出对其他车流的阻滞影响;且在低中高3种不同的流量需求下,改进模型计算结果与仿真输出的驶离流量基本相符,与原有模型相比精度更高;由此证实了改进模型的准确性.     

A New Congestion Prevention Policy for the Router

Introduction When too many packets are present in therouter, the performance of the router will degradeand the network will have congestion. Congestioncan be cause by several factors. If all of the suddenstreams of packets begin arriving on three or fourinput lines and all packet need the router to dealwith, a queue will build up. Slow processors canalso cause congestion. If the router’s CPU is slowat performing the bookkeeping tasks required ofthem, queues can build up even though there is…     

Queue length estimation at metered freeway‐to‐freeway connectors

Freeway‐to‐freeway connector metering is a cost‐effective and proven freeway management strategy for relieving recurrent congestion. However, one of the critical challenges in design and operation of freeway‐to‐freeway connector metering is the lack of up‐to‐date queue storage length design guidance. In this study, it was found that ramp queue is dynamically related to the metering rate, on‐ramp demand, and traffic flow arrival pattern. Hence, simply using an average demand cannot provide accurate queue length estimation and is also not suitable for queue storage design where the maximum or a percentile queue length is generally used. A mesoscopic queue length simulation model was developed based on the input–output method for estimating queue lengths under various demand‐to‐capacity ratio scenarios. Simulation results indicate that for under‐saturated situations, the ramp queue may exist temporally due to the random short‐term surge of traffic arrivals, and the exponential function could best capture the relationship between queue length and demand‐to‐capacity ratio. For over‐saturated situations, the ramp queue tends to prolong linearly with the demand‐to‐capacity ratio. Based on the simulation, it was recommended that queue storage length be designed as 4.3% of on‐ramp demand when demand is lower than 1200 vph or 2.3% when demand is between 1200 and 2400 vph. Copyright © 2016 John Wiley & Sons, Ltd.