基于微观动力学参数的高速公路复杂特征路段事故风险识别方法

高速公路是我国道路交通的命脉，已达14.3万公里，但目前每百公里致死率是其它等级道路的3倍多，且复合曲线、隧道、沿江桥等复杂特征路段具有事故多样性、集聚性的特点，而对其从微观动力学参数角度开展事故风险识别的系统研究较缺乏。鉴于此，项目运用自然驾驶、模拟驾驶、计算机仿真和数值分析相结合等手段，系统研究事故风险机理在微观动力学参数—中观（整车运行状态和驾驶行为参数）—宏观（交通和自然环境参数）之间的传递规律，提取基于微观动力学参数的复合型（耦合多维度因素影响）事故风险替代指标；加载事故风险类型和严重程度，考虑数据异质性和内生性，构建理想和实际环境下的高速公路复杂特征路段事故风险预测模型；探索以微观动力学参数为表征的行车实时风险预警与控制。通过对典型高速公路的事故风险识别分析，逐渐形成一套以微观动力学参数为核心的交通事故风险识别方法，为道路安全设计、行车安全风险预警与控制提供理论和方法支持。

The highway is the lifeblood of road traffic, which has reached 143,000 kilometers in China, but the current fatal rate per 100 kilometers is more than three times that of other grades. Although the segments with complex characteristics (such as composite alignment, tunnels, bridges along the river, etc) have feathers of crash diversity and agglomeration, systematic research on crash risk identification from the perspective of microscopic dynamic parameters is scarce. In view of this, this project will use a combination of naturalistic driving, simulated driving, computer simulation and numerical analysis to systematically study the transfer law of crash risk between micro-dynamic parameters, meso-factors (vehicle running state parameters and driving behavior parameters) and macro-parameters (traffic environment parameters and natural environment parameters). Then the substitution indexes for crash risk, coupling the influence of multi-dimensional factors, can be extracted based on micro-dynamic parameters. After loading crash risk types and severity, the crash risk prediction model which considers heterogeneity and endogeneity will be constructed for highway segments with complex characteristics under ideal and actual environments. Further the real-time driving risk early warning and control characterized by microscopic dynamic parameters will be explored. Through the analysis of crash risk identification of typical highways, a set of traffic crash risk identification methods with micro dynamic parameters as the core has been gradually formed, that can provide theoretical and methodological support for road safety design and traffic safety risk warning and control.