基于空间分析技术的交通事故多发点鉴别及成因分析

发布时间：2018-05-27 08:15

本文选题：空间分析 + 事故多发点鉴别　；参考：《清华大学》2014年硕士论文

【摘要】：近年来,交通事故已经成为世界范围内的公害。在我国随着社会经济的快速发展,汽车保有量急剧增长,交通事故也呈现多发的态势,道路交通安全问题已经成为交通发展中迫切需要解决的问题。道路交通安全的管理的重要基础之一是鉴别出道路交通事故的多发点(路段、区域),分析其主要的成因要素,进而提出切实可行的交通治理措施,从而提升我国的道路交通安全管理水平。本研究采取理论分析与应用实践相结合的研究思路,从理论、方法、技术、应用等层面应用空间分析技术重点研究了事故多发点的鉴别、特征及主要影响因素等内容。论文首先探讨了国内外学者在交通事故多发点鉴别与成因分析,以及空间分析方法在其它专业领域的应用成果,并梳理了现有研究存在的问题和不足;针对本研究所要使用的数据,通过数据清洗技术对“不完整的、错误的、重复的”三大类“事故脏数据”进行筛选和整理,避免因为事故数据的不完备性影响事故多发点鉴别和成因分析的准确性,并提出采取基于关系数据库的多维数据模型进行有序、层次性存储。针对本文的核心研究内容,首先结合事故多发点鉴别及成因分析的需求,介绍了缓冲区分析、叠置分析和核密度聚类三种空间分析方法的原理及应用方法。之后结合道路交通事故多发点鉴别及成因分析需求,提出了相应的改进三种空间分析技术的应用方法,包括“变长半径的点缓冲区分析”和“叠置分析”相结合的方法以及“优化窗宽的核密度聚类”的方法。之后基于交通管理部门在杭州市区范围一段时期内采集到的交通事故原始数据,应用提出的三种方法进行了研究区域内道路交通事故的多发点的鉴别,进而分析了导致交通事故多发的主要成因,并对不同方法的鉴别、分析结果进行对比评价。研究结果表明,本文应用改进的空间分析方法进行道路交通事故多发点的鉴别与成因分析,能够有效地对事故数据进行管理和分析,为日益严峻的交通安全问题研究做出一定的探索。而且,将来可以继续在交通事故预测方面做进一步的探索、研究。
[Abstract]:In recent years, traffic accidents have become a worldwide public hazard. With the rapid development of social economy in our country, the number of vehicles has increased rapidly, and the traffic accidents have become more and more frequent. The problem of road traffic safety has become an urgent problem to be solved in the development of traffic. One of the important bases of road traffic safety management is to identify the traffic accident prone areas (road sections, regions), analyze the main factors of their causes, and then put forward feasible traffic control measures. In order to improve our country's road traffic safety management level. This research adopts the research thought of combining the theory analysis with the application practice, from the theory, the method, the technology, the application and so on, applies the spatial analysis technology to study emphatically the identification, the characteristic and the main influence factor and so on of the accident frequent spot. Firstly, this paper discusses the identification and cause analysis of traffic accident prone sites by domestic and foreign scholars, as well as the application results of spatial analysis methods in other professional fields, and combs the existing problems and shortcomings of the existing research. In view of the data to be used in this study, the "incomplete, wrong and repeated" three categories of "accident dirty data" are screened and sorted by data cleaning technology. It is avoided that the incompleteness of accident data will affect the accuracy of identification and cause analysis of accident prone points, and a multidimensional data model based on relational database is proposed for orderly and hierarchical storage. According to the core contents of this paper, the principle and application of three spatial analysis methods, buffer analysis, overlay analysis and kernel density clustering, are introduced. Then combined with the needs of identification and cause analysis of road traffic accidents, the paper puts forward the application methods of improving three kinds of spatial analysis technology. It includes the combination of "point buffer analysis with variable radius" and "overlay analysis" and the method of "optimizing kernel density clustering with window width". Then, based on the original traffic accident data collected by the traffic management department in Hangzhou for a period of time, three methods are used to identify the road traffic accidents in the study area. The main causes of traffic accidents are analyzed, and the results are compared and evaluated. The results show that the improved spatial analysis method can be used to identify and analyze the traffic accident prone points, which can effectively manage and analyze the accident data. For the increasingly serious traffic safety problems to make a certain exploration. Furthermore, further exploration and research on traffic accident prediction can be made in the future.
【学位授予单位】：清华大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U491

【相似文献】