混合交通环境下车速离散时空演化及对交通安全的影响机理研究

As a comprehensive term of speed, speed dispersion will greatly impact on traffic safety performance, due to a combination of "human beings, vehicles, roads and traffic environment". Taking the township sections of arterial highways as examples, the common definition, distribution characteristics and laws of speed dispersion are studied under mixed traffic, considering the characteristic of mixed traffic in China. Then, the correlations and interactions between speed dispersion and high-dimensional factors under mixed traffic, based on data mining. With these correlations and relationships, the multilayer structural model and maps of speed dispersion temporal-spatial evolution are designed under mixed traffic. After analyzing the mechanisminfluencing factors of speed dispersion temporal-spatial evolution, the decision-makingprediction model for driving behaviorsof traffic conflicts is designed under speed dispersion. Meanwhile, with the studies on correlations among driving behaviors,between traffic conflicts and crash risk, the formation mechanism of traffic crash risk is explored under speed dispersion. Finally, a comprehensive model to assess traffic crash risk is designed to essentially explain the effects of speed dispersion on traffic safety. The study will provide a new idea or solution for speed dispersion evolution mechanism and traffic safety. And also, based on the study, much more reasonable and scientific measurements and methods can be designed for speed control and traffic hazard identification.

车速离散作为"人-车-道路-环境"共同作用下车速的总体表现，是影响交通安全的重要因素之一。项目以干线公路城镇段为对象，结合我国道路混合交通环境的基本特点，研究混合交通环境下车速离散程度的定义方法，探寻车速离散的基本特征与规律；基于数据挖掘技术，探索混合交通环境下高维度影响因素与车速离散的内在联系；构建多层次的车速离散结构模型及其时空演化图谱。在车速离散影响因素解析的基础上，借助交通冲突技术,构建车速离散背景下交通冲突预测模型，通过对交通冲突与事故风险之间的关联研究，揭示车速离散背景下交通事故风险产生机理，并开发交通安全风险评估模型，从本质上探寻车速离散对交通安全的影响作用。以期为系统认识混合交通环境下车速离散的形成及时空演化机理提供新方法，为研究车速离散与交通安全的关系提供新思路，为道路交通管理部门科学制定车速控制措施及危险源辨识提供理论依据。

车速离散作为“人-车-道路-环境”共同作用下车速的总体表现，是影响交通安全的重要因素。本课题以干线公路城镇段为研究对象，界定了其类型划分标准，提出了适用于混合交通环境下的断面及路段车速离散描述方法；基于无人机航拍数据，从路段及断面车速离散性两方面分析了车速离散的分布特征；基于因子分析法解析车速离散影响因素，并构建分层致因模型阐述其演化机理；构建了基于车速离散特征因素的交通冲突预测模型，并基于故障树（FTA）方法阐述了车速离散的严重交通冲突风险产生机理；以FTA模型为拓扑结构参照，构建了基于贝叶斯网络的安全风险评估模型，并验证了模型的可靠性；基于课题成果，提出了降低干线公路城镇路段车速离散性的处置对策。课题成果可为道路交通管理部门科学制定车速控制措施及危险源辨识提供理论依据。