基于网络核密度估计城市路网事故黑点鉴别研究

发布时间：2018-04-29 22:11

本文选题：交通安全 + 事故黑点鉴别　；参考：《东南大学》2015年硕士论文

【摘要】：已有统计数据显示,交通事故已成为危害人类健康和社会发展的重大问题,也是我国人群伤害死亡的第一位原因。而事故黑点路段因其事故易发,事故频发等特点,在道路安全改善项目中是需要优先整治的目标对象。然而已有的事故黑点鉴别方法对事故数据完整性要求较高,在实际工程应用中实施困难并且鉴别结果多是以文字和表格形式展现,可视性较差。因此本文应用网络核密度估计方法对城市路网事故黑点进行鉴别研究。核密度估计方法是空间分析理论中一种非参数估计理论,该方法不对样本数据分布进行事先假设,只需要获得事故发生的空间位置,并且可以基于GIS平台对鉴别结果进行多角度可视化展现。本文首先对核密度估计中的两个重要参数——窗宽和核函数对于核密度估计结果影响进行了讨论,接着对核密度估计方法应用于城市路网事故黑点鉴别中的问题作出简要说明,并对核密度估计方法做出修正以适应城市道路网络约束下的事故黑点路段鉴别要求,之后介绍了网络核密度估计方法应用于城市路网事故黑点鉴别的简要流程。紧接着本文对修正后的网络核密度估计方法应用于城市路网事故黑点鉴别的结果进行有效性检验,检验结果表明相比于已有传统的鉴别方法,网络核密度估计方法鉴别结果有效性更高,可以被考虑在工程应用中加以推广。本文之后应用ArcGIS Desktop软件,满足基于网络核密度估计城市路网事故黑点鉴别结果可视化需求。研究内容包括建立道路交通事故数据库,对道路网络对象进行单元划分,计算单元节点的网络核密度值,以及基于单元网络核密度值大小排序的事故黑点路段鉴别。本文研究的基于网络核密度事故城市路网事故黑点鉴别方法,其研究成果将会给城市道路安全治理中提供技术依据。
[Abstract]:Statistics have shown that traffic accidents have become a major problem that endangers human health and social development, and are also the first cause of death among people in China. Because of the characteristics of accident prone and frequent accidents, accident blackspot section is the target of priority treatment in road safety improvement project. However, the existing methods of accident blackspot identification require high integrity of accident data. It is difficult to implement in practical engineering applications and the identification results are mostly presented in the form of words and tables, and the visibility is poor. Therefore, the network kernel density estimation method is used to identify urban road network accident black spots. The kernel density estimation method is a non-parametric estimation theory in spatial analysis theory. The method does not presuppose the distribution of sample data and only needs to obtain the spatial location of the accident. And the discriminant result can be visualized based on GIS platform. In this paper, the influence of two important parameters in kernel density estimation, namely window width and kernel function, on the result of kernel density estimation is discussed, and then the application of kernel density estimation method to the identification of blackspots in urban road network accidents is briefly explained. The kernel density estimation method is modified to meet the identification requirements of accident blackspot sections under the constraints of urban road network. Then the brief flow of the application of network kernel density estimation method to the identification of urban road network accident blackspots is introduced. Then this paper verifies the validity of the modified kernel density estimation method applied to the blackspot identification of urban road network accidents, and the results show that compared with the existing traditional identification methods, The network kernel density estimation method is more effective and can be used in engineering applications. In this paper, ArcGIS Desktop software is applied to meet the visualization requirements of blackspot identification results of urban road network based on network kernel density estimation. The research contents include setting up the road traffic accident database, dividing the road network objects, calculating the network core density value of the unit node, and identifying the accident blackspot section based on the order of the unit network core density value. The black spot identification method of urban road network accidents based on network nuclear density is studied in this paper. The research results will provide technical basis for urban road safety management.
【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U491.31

【参考文献】