Early Detection of a Traffic Flow Breakdown in the Freeway Based on Dynamical Network Markers
Early Detection of a Traffic Flow Breakdown in the Freeway Based on Dynamical Network Markers
复制标题
DOI：
10.1007/s13177-019-00210-4
复制
发表时间：
2020-09-01
影响因子：
2.1
通讯作者：
Aihara, Kazuyuki
中科院分区：
文献类型：
--
作者：
Kamal, Md Abdus Samad;Oku, Makito;Aihara, Kazuyuki
本文提出了一种基于动态网络标记（DNM）的新型无模型方法，用于在联网车辆环境下与入口匝道相连的高速公路的背景下检测早期过渡阶段的交通流故障。在这种方法中，经常在每条车道上的几个单元或路段观察车辆状态。通过处理观测到的数据，分析细胞之间的标准差和相关系数，以确定主导细胞，即在转变过程中受影响最大的细胞。最后，将主导小区的相关系数的标准差和绝对值组合起来形成标量警告信号，在流量处于临界状态时提供非常强烈的指示。通过对高速公路交通进行仿真来评估所提出的方法，其流量受到匝道合流车辆的干扰。
This paper presents a novel model-free method based on the dynamical network markers (DNM) to detect the traffic flow breakdown at an early transition stage in the context of the freeway connected with an on-ramp under a connected vehicle environment. In this method, the vehicle states are frequently observed at several cells or segments on each lane. By processing the observed data, the standard deviations and the correlation coefficients among the cells are analyzed to determine the dominant cells, the ones that are mostly influenced during the transition. Finally, the standard deviations and absolute values of the correlation coefficients of the dominant cells are combined to form a scalar warning signal, which provides a very strong indication when the traffic is at the critical state. The proposed method is evaluated through simulation on freeway traffic, whose flows are disturbed by the on-ramp merging vehicles.