文章编号： 1672-5328（2026）02-0078-09

张建波1, 2，孙建平1, 2，张溪1, 2，张一鸣1, 2
(1. 北京交通发展研究院，北京100073；2.城市交通运行仿真与决策支持北京市重点实验室，北京100073)

摘要： 通过挖掘机动车风险驾驶行为数据，建立街道尺度的道路交通安全风险影响因素辨识方法， 旨在实现对超(特)大城市道路交通安全风险的精细化解析与量化评价。首先，基于机动车行驶轨迹 数据，设计一种考虑加速度变化强度与连续性的风险驾驶行为诊断方法，进而提取机动车急加速与 急减速行为特征；其次，结合道路建成环境与交通运行特征，构建面向街道尺度的道路交通安全影 响因素辨识方法，利用回归分析解析影响区域道路交通安全的环境因素；最后，提出街道级道路交 通安全风险指数评价方法，用于量化和追踪区域道路交通安全风险水平的动态变化。实证分析结果 表明：机动车风险驾驶行为频率分布与交通事故频率分布之间存在显著正相关性，且二者呈现相似 的双峰分布模式；车站里程密度、节点里程密度、严重拥堵里程比和主干路行驶速度标准差等因 素，是影响街道尺度下区域道路交通安全风险的关键因素。

关键词： 道路交通安全；风险驾驶行为；影响因素；回归分析；北京市

中图分类号： U491.3

文献标识码：A

Identification of Influencing Factors of Road Traffic Safety Based on Risky Driving Behaviors: A Case Study of Beijing

Zhang Jianbo1, 2, Sun Jianping1, 2, Zhang Xi1, 2, Zhang Yiming1, 2
(1. Beijing Transport Institute, Beijing 100073, China; 2. Beijing Key Laboratory of Urban Transport Simulation and Decision Making Support , Beijing 100073, China)

Abstract： By mining data on risky driving behaviors of motor vehicles, this paper establishes a street-scale methodology for identifying the influencing factors of road traffic safety risks, with the aim of enabling refined analysis and quantitative assessment of road traffic safety risks in megacities and super-large cities. First, based on vehicle trajectory data, a diagnostic method for risky driving behaviors is designed, which accounts for both the intensity and continuity of acceleration changes, thereby extracting characteristics of rapid acceleration and rapid deceleration behaviors of motor vehicles. Second, by integrating the road built environment and traffic operation characteristics, a street-scale identification method for identifying the influencing factors of road traffic safety is established, and regression analysis is utilized to analyze the environmental factors affecting regional road traffic safety. Finally, a street-level road traffic safety risk index evaluation method is proposed to quantify and track the dynamics changes in regional traffic safety risk levels. Empirical results indicate a significant positive correlation between the frequency distribution of risky driving behaviors and that of traffic accidents, with both exhibiting a similar bimodal distribution pattern. Key factors affecting regional road traffic safety risks at the street scale include station mileage density, node mileage density, proportion of severely congested road mileage, and the standard deviation of operating speeds on arterial roads.

Keywords： road traffic safety; risky driving behavior; influencing factors; regression analysis; Beijing