基于历史数据的高速公路交通事故严重程度分析

Analysis of Highway Traffic Accident Severity Based on Historical Data

摘要: 为深入探讨高速公路交通事故严重性的内在机制及其影响因素，本文构建了一个基于“人–车–路–环境–事故属性”五维框架的事故严重程度影响因素指标体系。首先，利用Pearson卡方检验筛选并确立事故影响因素及其之间的相关性；然后，将事故清理时间和占用车道数作为衡量事故严重程度的关键指标；最后，结合风险矩阵法(Risk Matrix Method, RM)对事故等级进行划分，将事故严重程度作为因变量，将事故影响因素作为自变量，构建有序Logit模型。论文研究发现：受伤人数、死亡人数、车辆数量、影响车道数和处置时间增加与事故严重性概率上升相关，而晴朗天气和非弯道事故情境则与降低严重性风险相关，为事故管理和预防提供了科学支持。

Abstract: To deeply explore the internal mechanism of the severity of freeway traffic accidents and their influencing factors, this study constructs an indicator system for the factors affecting accident severity based on a five-dimensional framework of “human-vehicle-road-environment-accident attribute”. Firstly, the Pearson chi-square test is used to screen and identify the accident-influencing factors and the correlations among them. Secondly, accident clearance time and the number of occupied lanes are taken as the key indicators to measure accident severity. Finally, combined with the Risk Matrix Method (RM) to classify accident levels, an ordered Logit model is established by taking accident severity as the dependent variable and accident-influencing factors as independent variables. The results of this study show that increases in the number of injured persons, the number of deaths, the number of vehicles involved, the number of affected lanes, and disposal time are associated with a higher probability of accident severity. In contrast, sunny weather and non-curved road accident scenarios are associated with a reduced risk of severity. These findings provide scientific support for accident management and prevention.