信号控制交叉口排队头车搜索算法

以信号控制交叉的口单条进口车道为研究对象，设计了一种基于实时检测数据的排队头车搜索算法．建立了由3辆车构成的期望排队头车集合，在停车线后布设上游检测器和停车线检测器，将信号周期时间划分为3个时间窗，用以采集车辆到达信息．通过分析绿灯末期进口车道的车辆运行特性，建立了若干逻辑条件．每个信号周期的红灯启亮时刻，根据逻辑条件的判断结果搜索该车道的排队头车．仿真测试结果显示，算法的正确率达到96．5％．由于检测器的布设位置、排队头车的车速阈值可以根据现实情况进行校正，使得算法具有较强的灵活性和适用性．

A Lead Queued Vehicle Determination Algorithm for Signalized Intersection

This paper focuses on the lead queued vehicle on a single approach of signalized intersection and presents a Lead Queued Vehicle Determination Algorithm on the basis of real-time detector data. It establishes an Expected Lead Queued Vehicle Set comprising three candidate vehicles. Two advanced detectors, i.e. upstream detector and stop line detector, are placed behind the stop line to collect vehicle arrival information within three time windows. A set of logic conditions is created on the basis of kinematics characteristic analysis of vehicle movement near the stop line during late part of green duration. For each signal cycle, according to the judgments of the logic conditions, the lead queued vehicle is determined when red light turns on. Simulation tests conducted by VISSIM and VisVAP indicate that the proposed algorithm is capable of identifying the actual lead queued vehicle with an accuracy of 96.5 %. Since the detectors location and the threshold speed value can be calibrated in the field, it provides the algorithm a wider flexibility in applications.