高速公路交通事故数据分析方法研究  

Study on Data Analysis Method of Expressway Traffic Accidents

作　　者：彭翔 宋灿灿[2] 王维利[3] 康可心[3] PENG Xiang;SONG Cancan;WANG Weili;KANG Kexin

机构地区：[1]同济大学建筑设计研究院(集团)有限公司,上海市200092 [2]同济大学道路与交通工程教育部重点实验室,上海市201804 [3]贵州省交通规划勘察设计研究院股份有限公司,贵州贵阳550081

出　　处：《城市道桥与防洪》2020年第5期225-229,M0023,共6页Urban Roads Bridges & Flood Control

摘　　要：高速公路交通事故数据对管理部门提升道路交通安全具有重要意义。为研究贵州省某两条高速公路历史交通事故数据分布规律与事故发展趋势,首先利用邻近度与关联性分析方法,完善事故数据;然后分析道路特征对交通安全的影响,划分连续下坡路段、隧道路段单元范围;最后对路段单元进一步划分为区块,建立不同区块范围内的事故概率与区块位置的预测模型,其中连续下坡路段后半段符合线形关系,隧道进出口段符合二次函数关系,并根据事故分布特征提出改善方案,进而辅助管理者掌握不同特征路段未来可能发生交通事故的路段范围以及改善的优先级。The expressway traffic accident data is of great significance for the management department to improve the road traffic safety.In order to study the historical traffic accident data distribution and accident development trend of two expressways in Guizhou Province,the proximity and correlation analysis method is used to perfect the accident data firstly,then the influence of road feature on the traffic safety is analyzed,the unit range of continuous downhill road section and tunnel road section is divided,and finally the unit of road section is further divided into the blocks to establish the prediction model of accident probability and block position in the range of different blocks,in which the second half section of continuous downhill road section conforms to the linear relation,and the tunnel entrance/exit section conforms to quadratic function relation.Based on the accident distribution features,the improvement scheme is proposed so that the managers are assisted to master the road section range of traffic accidents possibly to happen in the different features of road sections and to grasp the improvement priority.

关 键 词：交通事故 连续下坡 隧道 预测模型 

分 类 号：U491.3[交通运输工程—交通运输规划与管理]