高速公路桥梁路段交通安全风险评价及影响要素挖掘

Traffic Safety Evaluation and Feature Analysis of Freeway Bridge Sections

摘要:

为了深入探究影响高速公路桥梁路段风险的显著因素，论文应用交通秩序指数综合评估交通安全风险，并以此为因变量，研究桥梁路段交通流、道路属性及外部环境等要素和交通安全风险间的关系。在构建安全风险识别模型的基础上，通过个体条件期望图挖掘影响高速公路桥梁路段的安全风险关联因素。结果表明：相较于梯度提升决策树模型，随机森林模型对于交通安全风险识别的准确性更高；此外，在影响因素中，拥堵是影响桥梁路段安全风险的重要交通流因素；在能见度较低及不良天气情况下，交通风险较高；并且，跨河桥上、下游过渡段是高速公路桥梁路段中交通安全风险最高的区域。研究结果为高速公路桥梁路段的交通安全风险识别及影响要素挖掘提供了新的思路和方法，有助于交通管理部门精准实施靶向治理。

Abstract:

In order to explore in depth the significant factors affecting the traffic risk of bridge sections of freeway， this paper taking traffic order index （TOI） as dependent variable， and identifies the relationship between TOI and traffic characteristics， road attributes， and external environment. Based on the safety evaluation model， the Plots of individual conditional expectation （ICE） were applied to discover the contributing factors to traffic risk of bridge sections of freeway. The results indicate that compared with the gradient boosting decision tree model， the random forest（RF） models is more accurate for traffic safety risk evaluation. In addition， among the influencing factors， congestion is an important traffic flow factor affecting the safety risk of bridge sections. Under low visibility and adverse weather conditions， traffic risks are higher. The upstream and downstream of the river-crossing bridge are considered high-risk corridors in the bridge sections of the freeway. This paper can provide new ideas and methods for traffic safety risk evaluation and discovery of contributing factors to traffic risks of bridge sections， which is helpful for the traffic management department to accurately implement management.