基于IHSDM框架的高速公路交通事故预测模型

发布时间：2018-04-26 00:10

本文选题：高速公路 + 基础事故预测模型　；参考：《哈尔滨工业大学》2014年硕士论文

【摘要】：不良的线形条件是诱发高速公路交通事故的重要原因之一，分析高速公路线形与事故的关系，建立基于线形条件的事故预测模型，对于优化高速公路线形设计及减少事故发生有重要意义。美国联邦公路局开发的IHSDM事故预测模型由于建模思想可靠、实用性强等特点，得到了广泛推崇与研究。本文借鉴IHSDM事故预测模型的建模思想，利用我国高速公路交通事故及线形数据，建立了用于山岭区、重丘区、平原区高速公路的事故预测模型。首先将高速公路按同质法划分成几何线形上不可再分的预测单元；将收集到的交通事故数据及录像法获得的交通量数据对应到相应的预测单元上，据此建立了包含线形指标、交通事故、交通量等信息的山岭区、重丘区、平原区高速公路交通事故基础数据管理系统。分析了三种地形条件下高速公路平曲线路段和纵坡路段等单一线形指标路段以及弯坡组合路段等组合线形上事故率与线形指标的关系，鉴别出对事故有突出影响的线形指标，如平曲线半径及纵坡坡度等。在此基础上，界定了事故率最低时线形指标所处的范围，即理想的线形条件。分析理想线形上路段事故率与交通量的关系，建立了基于年平均日交通量（AADT）的高速公路基础事故预测模型；选取山岭区、重丘区高速公路平曲线路段、偏角路段、直线路段、纵坡路段、竖曲线路段，选取平原区高速公路直线路段，分析实际事故率与基础事故率的关系，确定了上述线形指标对基础事故的修正系数。整合基本事故预测模型与线形指标修正系数，得到了一套完整的用于山岭区、重丘区、平原区高速公路事故预测的模型。利用辽宁省沈丹高速公路、沈大高速公路、铁阜高速公路线形及事故数据，对模型进行验证。结果表明本文所建模型具有一定的精度。最后，，利用本文建立的事故预测模型对广东省潮惠高速公路进行事故预测。
[Abstract]:The bad linear condition is one of the important reasons to induce the expressway traffic accident. The relationship between the expressway alignment and the accident is analyzed, and the accident prediction model based on the linear condition is established. It is of great significance for optimizing the design of expressway alignment and reducing accidents. The IHSDM accident prediction model developed by the Federal Highway Administration of the United States has been widely respected and studied because of its reliable modeling idea and strong practicability. Based on the modeling idea of IHSDM accident prediction model and the traffic accident and linear data of expressway in China, the accident prediction model for mountain area, heavy hill area and plain area is established in this paper. Firstly, the highway is divided into geometric linear prediction units according to homogeneous method, traffic accident data collected and traffic volume data obtained by video recording method are corresponding to the corresponding prediction units, and the linear index is established. Traffic accident, traffic volume and other information mountain area, heavy hill area, plain area expressway traffic accident basic data management system. In this paper, the relationship between the accident rate and the linear index is analyzed under three kinds of terrain conditions, such as single linear index section such as plain curve section and longitudinal slope section of expressway, and composite road section of curved slope, and the linear index which has prominent influence on the accident is identified. Such as the radius of the plane curve and the slope of the longitudinal slope and so on. On this basis, the range of linear index at the lowest accident rate is defined, that is, the ideal linear condition. Based on the analysis of the relationship between accident rate and traffic volume on the ideal line, the prediction model of expressway basic accident based on the annual average daily traffic volume (AADT) is established, and the highway plane curve section, deflection section and straight line section are selected in mountain area and heavy hill area. The relationship between the actual accident rate and the basic accident rate is analyzed, and the correction coefficient of the above linear index to the foundation accident is determined by analyzing the relationship between the actual accident rate and the basic accident rate, and selecting the straight section of the highway in the plain area as well as the vertical curve section. By integrating the basic accident prediction model with the linear index correction coefficient, a complete set of models for highway accident prediction in mountain area, heavy hill area and plain area is obtained. The model is verified by the alignment and accident data of Shendan Expressway, Shen-Da Expressway and Tiefu Expressway in Liaoning Province. The results show that the model has a certain accuracy. Finally, the accident prediction model established in this paper is used to predict the accident of Chaohui Expressway in Guangdong Province.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491.31

【参考文献】