摘要
长周期和倒计时是中国相对于发达国家城市道路交叉口信号控制方案的两个独有特征。基于实证数据分析两类信号控制特征对车辆排队消散特性的影响。共观测天津市两处典型的长周期倒计时信号控制交叉口的7条直行车道。实证数据表明,中国城市道路交叉口长周期倒计时信号控制下的排队消散过程大致可分为3个阶段:启动阶段、稳定阶段和上升阶段。通过有序样品聚类的方法确定3个阶段的分割点。这与HCM 2010等经典理论假设的排队消散特性明显不同,因此分析传统通行能力估计算法HCM 2010的适用性,并提出简单线性法、抛物线法和两段线法3种通行能力估计方法以减少传统方法估计中的误差。通过比较发现,抛物线与两段线拟合方法对通行能力的估计较为准确,均方根误差(RMSE)均小于30 pcu·h^(-1)。
Long cycle and duration countdown are two unique characteristics of signalized intersections in China by comparison to those developed countries. This paper empirically investigates the impacts of such two characteristics on queue discharge at signalized intersections. For this purpose, seven traffic lanes at two typical intersections with long cycle and duration countdown are selected as study sites in Tianjin. The results show that the discharge process of queuing vehicles can be divided into three stages: start-up stage,stable stage, and rising stage. The ordered samples clustering method is used to identify the three partitions in which the findings are distinguished from the conventional assumption of capacity estimation methodologies, namely HCM 2010, the adaptiveness of traditional methods are thus analyzed. Furthermore, PieceWise Linear Method, Quadratic Linear Method, and Simple Linear Method are proposed to implement a better capacity estimation. The comparison results suggest a relatively more precise solutions from PieceWise Linear Method and Quadratic Linear Method where the RMSE is less than 30 pcu/h.
出处
《城市交通》
北大核心
2016年第6期67-74,共8页
Urban Transport of China
关键词
城市交通
信号控制
排队消散特性
通行能力
有序样品聚类
长周期
倒计时
urban transportation
signal control
queue discharge characteristics
capacity
ordered samples clustering
long cycle
duration countdown