基于轨迹数据的单交叉口信号配时优化方法

Signal Timing of Intersection Based on Trajectory Data

随着城市化的发展,城市路网变得更加复杂,交通量日益增多。传统的数据采集方式如线圈、卡口存在铺设难度大、维护成本高、数据缺失等缺点。随着网络出租车的兴起,海量的轨迹数据被收集起来。本文整合了基于轨迹数据的交叉口排队模型和交叉口启动波模型,估计路口各相位排队长度,并应用交通波理论对交叉口各相位配时进行优化,建立了基于轨迹数据的单交叉口信号配时模型。该模型采用交叉口SCATS系统实际数据进行路口建模,经仿真验证,相较于原配时方案与Webster配时方法,该模型能够将交叉口延误分别降低61.37%和36.95%,验证了该模型的有效性。

With the development of urbanization,urban road network has become more complicated and traffic vol⁃ume is increasing.Traditional data collection methods such as coils and bayonet have disadvantages of difficult instal⁃lation,high maintenance cost,and lack of data.With the rise of online taxis,massive trajectory data can be collect⁃ed.This paper integrates an intersection queuing model and an intersection starting wave model based on trajectory da⁃ta to estimates the queue length of each phase of an intersection,and then uses traffic wave theory to optimize the tim⁃ing of each phase of the intersection.Therefore,this signal timing model for single intersection based on trajectory da⁃ta is developed.The model uses actual data in the intersection SCATS system for intersection modeling.Compared with the original timing scheme and Webster timing method,the model can reduce the intersection delay by 61.37%and 36.95%,respectively,which validates the effectiveness of the model.