基于滴滴轨迹数据的城市道路运行状态评价分析

发布时间：2018-08-07 08:30

【摘要】：随着城市化进程的不断加快,交通拥堵已成为城市面临的主要社会问题之一。交通拥堵会制约城市的发展,增加不必要的社会出行成本,引起城市经济的大量损失。因此实时、准确的判断道路运行状态以及长时间连续地监控道路运行状态的变化规律,对制定更加合理的交通道路管理策略具有显著意义。故本文基于以上背景,运用滴滴专快车轨迹数据,利用地图匹配技术,从点线面三个层次建立城市道路运行状态评价体系,使用C#编程语言,搭建城市交通运行状态实时监控系统。本文首先介绍所使用的滴滴车辆轨迹数据的结构与范围,并对原始数据进行预处理,其中包括轨迹坐标的转换和对坏点数据的剔除与插值。其次介绍了地图匹配的常用算法和本文所使用的垂线投影法与历史轨迹校核法,综合利用垂线投影法简单高效和历史轨迹校核法准确度高的特点,本文所采用的综合地图匹配算法具有较高匹配效率和较低匹配误差的优势。再次,根据滴滴车辆轨迹数据对一条城市道路的交通流特性进行具体分析,给出道路车流量、速度、密度的变化,以及三者的关系,并对比工作日与非工作日交通流特性的变化,以及早晚高峰对交通流特性的影响。然后,从点线面三个层次,划分不同道路等级,使用道路流量,对不同等级的道路赋予不同的权重,并以延误指数为评价指标,来综合评价路网的运行状态。其中,点层次是指单一路段的运行状态评价;线层次是指不同等级的道路运行状态评价;面层次是指一定区域范围内的路网整体运行状态评价。最后依据地图匹配方法和道路运行状态评价方法,在.Net平台上,通过C#编程语言,结合ArcEngine组件库,搭建出交通运行状态评价实时监控系统,对路网中所有道路进行实时监控,进而评价某一路段、某等级道路或某区域路网的交通运行状态。
[Abstract]:With the acceleration of urbanization, traffic congestion has become one of the main social problems faced by cities. Traffic jams will restrict the development of cities, increase unnecessary social travel costs, and cause a lot of losses in urban economy. Therefore the real-time and accurate judgment of the road running state and the continuous monitoring of the changing law of the road running state for a long time have significant significance to the formulation of more reasonable traffic road management strategy. Therefore, based on the above background, using Didi special express track data, using map matching technology, the evaluation system of urban road running state is established from three levels of point, line and surface, and the C # programming language is used. Set up the real-time monitoring system of urban traffic running state. This paper first introduces the structure and range of the Didi vehicle track data, and preprocesses the original data, including the transformation of trace coordinates and the elimination and interpolation of bad point data. Secondly, it introduces the common algorithms of map matching, the vertical projection method and the historical track checking method, which are used in this paper, and makes comprehensive use of the characteristics of the vertical projection method which is simple and efficient and the historical track checking method with high accuracy. The synthetic map matching algorithm adopted in this paper has the advantages of high matching efficiency and low matching error. Thirdly, according to Didi's vehicle track data, the characteristics of traffic flow on a city road are analyzed in detail, and the changes of traffic flow, speed, density and the relationship among them are given. The change of traffic flow characteristics between working day and non-working day and the influence of morning and evening rush hour on traffic flow characteristics are compared. Then, from three levels of point line, different road grade is divided, road flow is used, different weight is given to different grade road, and delay index is taken as evaluation index to evaluate the running state of road network synthetically. Among them, the point level refers to a single section of the operation state evaluation; line level refers to different levels of road operation state evaluation; plane level refers to a certain area within the overall operation of the road network state evaluation. Finally, according to map matching method and road running state evaluation method, on the platform of. Net, through C # programming language, combined with ArcEngine component library, a real time monitoring system of traffic running state evaluation is built, and all roads in the road network are monitored in real time. And then evaluate the traffic running state of a certain section, a certain grade road or a regional road network.
【学位授予单位】：北京交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491

【参考文献】