交通运输系统工程与信息

交通运输系统工程与信息 ›› 2021, Vol. 21 ›› Issue (2): 166-172.

• 系统工程理论与方法 • 上一篇    下一篇

基于风险矩阵的干线公路弯道路段交通冲突风险评估模型

房锐1，张琪1, 2，胡澄宇1, 3，田毕江1，覃文文*2   

  1. 1. 云南省交通规划设计研究院有限公司，陆地交通气象灾害防治技术国家工程实验室，昆明 650200； 2. 昆明理工大学，交通工程学院，昆明 650504；3. 同济大学，道路与交通工程教育部重点实验室，上海 201804
  • 收稿日期:2020-09-26 修回日期:2021-01-15 出版日期:2021-04-25 发布日期:2021-04-25
  • 作者简介:房锐(1967- )，男，云南曲靖人，教授级高级工程师。
  • 基金资助:

    国家重点研发计划/National Key Research and Development Program of China(2017YFC0803906)；国家自然科学基金/National Natural Science Foundation of China(52062024，52002161)。

Risk Assessment Model Based on Risk Matrix for Traffic Conflict on Arterial Highway Bend Section

FANG Rui1 , ZHANG Qi1, 2 , HU Cheng-yu1, 3 , TIAN Bi-jiang1 , QIN Wen-wen*2   

  1. 1. National Engineering Laboratory for Surface Transportation Weather Impacts Prevention, Broadvision Engineering Consultants Co. Ltd, Kunming 650200, China; 2. Faculty of Traffic Engineering, Kunming University of Science and Technology, Kunming 650504, China; 3. Key Laboratory of Road and Traffic Engineering, Ministry of Education, Tongji University, Shanghai 201804, China
  • Received:2020-09-26 Revised:2021-01-15 Online:2021-04-25 Published:2021-04-25

PDF

摘要：

为评估干线公路弯道路段(AHBS)交通冲突风险，基于交通冲突前5 min集计交通流数据、冲突路段道路线形特征和行车环境数据，分别建立交通冲突可能性及严重度评估指标体系。集成运用随机森林模型和多层次模糊综合评价法分别确定交通冲突可能性和严重度等级，进而采用风险矩阵法确定交通冲突风险等级。以云南省元双干线公路为例进行验证。结果表明：基于随机森林模型的交通冲突可能性预测准确率达到84.21%，所提模型能有效评估AHBS交通冲突风险，为干线公路交通事故治理提供理论依据。

关键词: 交通工程, 交通冲突风险评估模型, 风险矩阵, 干线公路弯道路段, 随机森林

Abstract:

In order to evaluate the traffic conflict risk on arterial highway bend section (AHBS), an index framework for assessing possibility and severity of traffic conflict was established based on five-minute aggregation traffic flow data before a traffic conflict, road alignment characteristics in conflict sections, and driving environment data. An integration of random forest model and multi- level fuzzy comprehensive evaluation were used to determine the probability and severity level of traffic conflict. The risk matrix method was used to determine the comprehensive risk level of traffic conflict. Finally, a verification analysis was conducted with an example of the Yuan Shuang arterial highway. The results show that, the probability accuracy of traffic conflict based on random forest model can reach to 84.21% . The proposed model can effectively evaluate the traffic conflict risk in AHBS, and a theoretical basis is provided to reduce traffic accidents on arterial highway.

Key words: traffic engineering, traffic conflict risk assessment model, risk matrix, arterial highway bend section, random forest

中图分类号: