高速公路出口匝道事故预测模型优选及弹性分析

发布时间：2019-06-09 18:25

【摘要】：为探索高速公路出口事故发生的关键诱因,依托美国佛罗里达州24条高速公路上405个出口匝道的历史事故和道路交通数据,验证了出口匝道事故服从于对数正态分布.以匝道交通量、匝道长度和设计一致性(分别以平均半径、曲率变化率、运行速度差和运行速度变化率度量)为解释变量,以2004—2006年间事故数为因变量,建立了4个泊松对数正态事故预测模型,其中以速度变化率表征设计一致性的事故预测模型具有最好的拟合度.基于最优拟合度模型的弹性分析表明,运行速度变化率及匝道长度为关键因素;基于安全考虑,出口匝道速度变化率宜控制在20%以内,出口匝道极限最小长度不宜小于200 m,一般最小长度不宜小于400 m.
[Abstract]:In order to explore the key inducement of highway exit accident, based on the historical accidents and road traffic data of 405 exit ramp on 24 highways in Florida, it is verified that the exit ramp accident obeys logarithmic normal distribution. Taking ramp traffic volume, ramp length and design consistency (measured by average radius, curvature change rate, running speed difference and running speed change rate, respectively) as explanatory variables, and taking the number of accidents in 2004 / 2006 as dependent variables, Four Poisson lognormal accident prediction models are established, among which the accident prediction model, which is characterized by speed change rate, has the best fitting degree. The elastic analysis based on the optimal fitting degree model shows that the change rate of running speed and ramp length are the key factors. Based on safety considerations, the speed change rate of exit ramp should be controlled within 20%, the minimum length of exit ramp should not be less than 200 m, and the general minimum length should not be less than 400 m.
【作者单位】：南京理工大学自动化学院;杜伊斯堡-埃森大学东亚研究学院;自由大学城市地理科学研究所;
【基金】：国家自然科学基金青年基金资助项目(51208261) 教育部人文社会科学青年基金资助项目(12YCZH062) 中央高校基本科研业务费专项资金资助项目(30920140132033) 美国佛罗里达州交通部资助项目(BD54438)
【分类号】：U491.31

【相似文献】