交通运输系统工程与信息 ›› 2019, Vol. 19 ›› Issue (2): 202-208.

• 系统工程理论与方法 • 上一篇    下一篇

基于贝叶斯网的公路路侧事故风险评价

蔡明 1,程国柱 2,张素禄 3,王波 1,周侃*1   

  1. 1. 深圳市市政设计研究院有限公司,广东 深圳 518029;2. 东北林业大学 交通学院,哈尔滨 150040; 3. 哈尔滨工业大学 交通科学与工程学院,哈尔滨 150040
  • 收稿日期:2018-11-07 修回日期:2019-01-13 出版日期:2019-04-25 发布日期:2019-04-25
  • 作者简介:蔡明(1961-),男,福建安溪人,教授级高级工程师.
  • 基金资助:

    国家自然科学基金/National Natural Science Foundation of China(51778063).

Risk Assessment of Highway Roadside Accidents Based on Bayesian Network

CAI Ming1, CHENG Guo-zhu2, ZHANG Su-lu3, WANG Bo1, ZHOU Kan1   

  1. 1. Shenzhen Municipal Design & Research Institute Co., Ltd, Shenzhen 518029, Guangdong, China; 2. College of Traffic, Northeast Forestry University, Harbin 150040, China; 3. School of Transportation Science and Engineering, Harbin Institute of Technology, Harbin 150040, China
  • Received:2018-11-07 Revised:2019-01-13 Online:2019-04-25 Published:2019-04-25

摘要:

为了根据路侧危险情况采取针对性的措施,急需开展公路路侧事故风险评价研究. 基于事故模拟再现分析,根据车辆驶入路侧后的事故形态,将路侧事故严重程度划分为 4个等级,对不同等级路侧事故的致因因素进行研究,给出驶入车速、边坡坡度与边坡高度阈值.构建了路侧事故风险评价的贝叶斯网,基于路侧事故致因因素阈值,针对单因素、双因素与三因素分别给出了不同等级路侧事故发生概率的计算方法.选取典型路侧事故案例进行分析,贝叶斯网的计算结果显示发生 2次及以下翻车事故的概率为 0.929,与路侧事故模拟试验结果相符,证明了本文提出的公路路侧事故风险评价方法的准确性.

关键词: 交通工程, 公路路侧事故, 风险评价, 贝叶斯网, 事故模拟

Abstract:

It is necessary to take targeted measures according to roadside dangerous conditions. Therefore, it is urgent to study roadside accident risk evaluation. Based on accident simulation analysis, roadside accident severity was divided into four grades. Thresholds of driving speed, grade and height of side slope were given according to study on different roadside accident severity contributing factors. Bayesian network of roadside accident risk assessment was constructed. Based on thresholds of roadside accident contributing factors, calculation methods corresponding to single- factor, two- factor and three- factor were presented for probability of different roadside accidents severity. Lastly, case analysis was carried out by selecting typical roadside accident. Bayesian network calculating results show that the probability of no more than two turnovers is 0.929 and it is in conformity with accident simulation results, which proves the accuracy of the method.

Key words: traffic engineering, highway roadside accident, risk evaluation, Bayesian network, accident simulation

中图分类号: