山区高速隧道交通事故严重程度预测及特大事故决策规则提取  

Prediction of traffic accident severity and extraction of decision rules for extraordinarily serious accident in mountainous high-speed tunnels

作　　者：乔建刚 范颖蓉 陶瑞 王傑 QIAO Jian’gang;FAN Yingrong;TAO Rui;WANG Jie(School of Civil Engineering and Transportation,Hebei University of Technology,Tianjin 300401,China)

机构地区：[1]河北工业大学土木与交通学院,天津300401

出　　处：《中国安全生产科学技术》2024年第7期186-192,共7页Journal of Safety Science and Technology

基　　金：国家重点研发计划重点专项(2017YFC0805404);国家安全监管总局2017年安全生产重特大事故防治关键技术科技项目(hebei-0009-2017AQ)。

摘　　要：为有效避免隧道段发生人员伤亡严重的交通事故,选取2013—2023年我国国内(不含港澳台)发生的交通事故数据进行统计分析,结合事故严重程度和时空分布情况筛选出14个影响因素;采用随机森林模型构建山区高速隧道段交通事故严重程度预测模型,对比分析有序Logit模型和BP神经网络模型与所构建的模型预测精度;基于规则重要性对随机森林中“特大事故”决策规则进行提取。研究结果表明:随机森林模型对于事故严重程度的预测结果较优,决策规则揭示人员伤亡严重时的影响因素组合。研究结果可为针对事故严重程度影响机理提出改进意见提供参考。In order to effectively avoid the traffic accidents in tunnel section with serious casualties,the historical traffic accident data from 2013 to 2023 in China were selected for statistical analysis,and 14 influencing factors were selected combined with the severity and spatiotemporal distribution of accidents.The random forest model was used to construct a prediction model for the severity of traffic accidents in mountainous high-speed tunnel section,and the prediction accuracy of this model was compared with those of the ordered Logit model and BP neural network model.The decision rules for“extraordinarily serious accident”in random forest were extracted based on the importance of rules.The results show that the random forest model has better prediction results for the severity of accidents,and the decision rules reveal the combination of influencing factors when the casualties are severe.The research results can provide reference for improving the influence mechanism of accident severity.

关 键 词：隧道 随机森林模型 决策规则 事故 严重程度 

分 类 号：X951[环境科学与工程—安全科学]