交叉口信号控制方案评价指标动态估计模型

对评价指标在周期内的实时变化进行动态估计是交叉口交通信号动态控制的关键。应用计算机仿真模拟方法以仿真步长为间隔建立了停车延误、停车次数、排队车辆数等常用控制目标与信号灯状态及持续时间之间的动态迭代估计模型，机动车流在路段上的运行特征用CTM描述，在交叉口上的演化规律用微观交通状态描述。以单个进口道的直行车流为研究对象，对经典延误模型进行对比分析，发现随着饱和度增大，各个模型的估计偏差有逐渐增大的趋势，基于仿真步长的迭代估计累积结果基本上处于经典延误模型计算结果的变化范围内。对于单个交叉口而言，迭代算法的时间复杂度为O(n)。

Dynamic Estimation Model on Performance Indexes of Signalized Intersection

It is necessary for dynamic traffic signal control on urban road network to dynamically estimate real-time changes of performance indices in one cycle. Simulation method was used for real-time estimation of performance indices at signalized intersection. The dynamic iterative relationship between vehicle delay, stops, and queues at intersection and signal status was modeled based on simulation step. Characteristics of traffic flow on link were simulated by Cell Transmission Model (CTM) while those of evolution at intersection described by Micro Traffic State. Delay incurred by straightforward vehicle flow on an approach of a signalized intersection was researched. Results showed that variance among several classical delay models will increase as degree of saturation increases. Analyses showed that estimation result based on simulation step is mainly fallen into the range of variance among these classical delay models. For one intersection, time complexity of the iterative algorithm is O(n).