密空聚类式交通事故多发路段智能鉴别研究

Research on Intelligent Identification of Accident-prone Sections Based on Density Space Clustering

以高速公路交通事故记录数据为基础,将聚类算法与核密度估计联动,依据我国对事故黑点的鉴别标准,进行交通事故密度聚类分析,参考我国对事故数据进行分级处理的事故标准,基于聚类结果计算交通事故黑点权重,进行交通事故空间数据核密度估计,并采用GIS技术可视化,运用空间统计方法智能鉴别多发事故路段,对高速公路警示标志设置和路安部门监管重点提供参考依据。通过对辽宁省某区域内的三条高速路段进行交通事故多发路段鉴别,智能鉴别能够有效应对不同交通事故点分布特点,避免简易事故的影响,清晰准确找到事故多发路段的位置。

Based on the actual traffic accident data of highway,the clustering algorithm is linked with the kernel density estimation,and the traffic accident density clustering analysis is carried out according to the identification standard of the traffic accident black spot in China.With reference to the accident standard of China for grading the accident data,the weight of the traffic accident black spot is calculated based on the clustering results,and the kernel density of the traffic accident spatial data is estimated,and the GIS technology is used for visualization,and spatial statistical methods are used to intelligently identify accident-prone sections,which provides a reference for the setting of warning signs on highways and the supervision of road safety departments.Through the identification of three highway sections with frequent traffic accidents in a certain area of Liaoning Province,intelligent identification can effectively deal with the distribution characteristics of different accident-prone sections,avoid the impact of simple accidents,and clearly and accurately find the location of accident-prone sections.