基于导航数据的交叉口进口道安全风险评估及诊断方法

Navigation-data-based Risk Evaluation Method at Intersection Entrance

摘要:

以导航软件采集的驾驶行为、空间位置、交通状况及用户量等大数据为基础，应用交通秩序指数评价道路安全性。通过分析与进口道相关的道路几何特征、交通执法设备以及信号控制设施的安全性特征，构建交叉口安全性结构方程模型，深入挖掘交叉口进口道风险致因。结果表明，交叉口各方向车道总数、信号相位，与进口有关的进口道车道数、调头车道数、横断面类型、中央分隔带类型、违法监控设备数量，都对交叉口安全性具有显著影响。拥堵状态对于交叉口安全性的影响最大，且在不同时段交叉口安全性存在较大差异。其次，信号控制设施与道路几何特征，交通执法设备对交叉口的安全影响最小。在改善交叉口进口道的安全性时，首要是鼓励错峰出行、改善拥堵状态，尽量采用带有全红时间的信号配时方案。道路设计过程中，在进口方向增加车道数量并设置硬性中央分隔带，尽量设置调头车道。提出一种基于导航数据的进口道风险与交叉口相关属性间量化关系的挖掘方法，进而实现交叉口进口道风险的评估与诊断，为其他类型道路及交通网全域风险防控和安全改造提供借鉴和参考。

Abstract:

The research adopted the traffic big data such as driving behavior，maps，traffic conditions and user volume collected by the navigation software to develop a traffic order index and evaluate the road safety risks.By analyzing the traffic safety features of road alignment，traffic control devices，traffic signals at intersection entrance，a structural equation model of intersection was build，the causes of risk at intersection entrance were analyzed in depth.The results show that the number of lanes in all directions，signal phase，the number of lanes related to the entrance，the number of U-turn lanes，the type of cross section，the type of median strip and the number of illegal behavior monitoring equipment have significant impact on traffic safety of the intersection.Congestion factors have the greatest impact on traffic safety of the intersection，then traffic signal factors and road alignment factors，and illegal behavior monitoring equipment have the least impact on traffic safety of the intersection.On improving the safety of intersection entrance， people are encourged to travel at off-peak hours to avoid the congestion，and all-red signals are proposed.In road designing， lanes should be increased in the entrance direction and median strip should be set，and the U-turn lanes should be set if the space allows.A safety evaluation method of intersection roads，facilities and signals based on driving behavior data collected by navigation software is proposed，which provides a reference for safety evaluation and cause analysis of other road types and transportation network.