应用神经网络动态估计信号交叉口饱和流率  

Dynamic estimation of saturation flow rate at signalized intersection used by neural network

作　　者：王益 荣建[3] 周晨静 高亚聪 罗薇 WANG Yi;RONG Jian;ZHOU Chen-jing;GAO Ya-cong;LUO Wei(School of Civil Engineering,Tsinghua University,Beijing 100084,China;Institute of Transportation Engineering,Tsinghua University,Beijing 100084,China;Beijing Key Laboratory of Traffic Engineering,Beijing University of Technology,Beijing 100124,China;School of Civil and Transportation Engineering,Beijing University of Civil Engineering and Architecture,Beijing 100044,China)

机构地区：[1]清华大学土木水利学院,北京100084 [2]清华大学交通研究所,北京100084 [3]北京工业大学北京市交通工程重点实验室,北京100124 [4]北京建筑大学土木与交通工程学院,北京100044

出　　处：《广西大学学报（自然科学版）》2021年第3期714-723,共10页Journal of Guangxi University（Natural Science Edition）

基　　金：国家自然科学基金资助项目(51708017);北京市博士后工作经费资助项目(2020-zz-089)。

摘　　要：为了实时掌握信号交叉口饱和流率动态变化规律和提升估算精度,构建了以神经网络为基础的饱和流率动态估计模型。通过对北京市典型信号交叉口3种场景(直行进口道、直行左转进口道、直行右转进口道)实测数据为研究对象,分析每种场景下交通流运行特征,确定影响饱和流率的关键因素,确定神经网络模型的输入输出参数,并对模型进行标定。最后与经典HCM方法进行对比。结果表明:不同场景下,神经网络模型估计精度均优于HCM方法;其估算误差分别为11.23%,7.02%,4.70%。提出的方法能够准确地动态估计饱和流率,成果可用于信号控制方案的实时调整与精细化的运行管理。In order to improve the accuracy of the saturated flow rate estimation at the signalized intersection,and timely grasp the dynamic change information of the saturation flow rates and the influence factors,a dynamic estimation method for saturated flow rate based on neural network was constructed.The measure data at three scenarios(through lanes,through-left lanes,through-right lanes)of signalized intersections in Beijing were taken as examples to validate the proposed method.Firstly,the traffic flow characteristics of the three scenarios and factors affecting the saturation flow rate were analyzed.Secondly,neural network models of the three scenarios were established and their hyperparameters were determined.Lastly,the proposed method was compared with the HCM method.The results show that neural network models have better accuracy than HCM models.With the increasing complexity of scenarios,the advantages of neural network models are highlighted.In the through lane,through-left lane and through-right lane scenarios,the estimated saturation flow rates used by the proposed method are 11.23%,7.02%,4.70%,respectively.The proposed method can estimate the saturation flow rate accurately and timely.and the results can be used for signal control schemes optimizing and subtle operation managing at signalized intersections.

关 键 词：信号交叉口 动态估计 神经网络 饱和流率 

分 类 号：U491[交通运输工程—交通运输规划与管理]