基于公交车大数据挖掘的城市路网自动生成与实时更新算法研究

发布时间：2018-03-29 03:36

本文选题：ITS　切入点：公交车数据　出处：《山东大学》2015年硕士论文

【摘要】：城市路网矢量地图是ITS (Intelligent Transportation System)的核心组成部分,在众多应用系统中都扮演着重要角色,如位置服务、路径规划、智能交通诱导系统等都离不开精确、完整的路网矢量地图。然而,随着城市经济的快速发展,路网结构日趋复杂而且变动频繁,给路网地图的维护带来更大挑战。目前,路网矢量地图大多依赖手工完成,成本高且更新周期较长,不易及时反映出路网结构的变动。因此,研究一种低成本且高效的城市路网地图自动生成方法有着极高的应用价值。近年来,越来越多的GPS车载设备被安装在了公交车、出租车上,这些浮动车分布于城市的各条道路上,通过GPS定位,实时上传着车辆的位置信息,大量的轨迹数据蕴藏着丰富的道路信息,而且这些数据具有精度高、道路覆盖范围广、采集成本低等优势。本文在深入分析公交车GPS数据特点的基础上,提出了一种基于公交车GPS数据自动生成城市路网地图的方法。论文首先对课题中涉及到的相关技术作了介绍,主要有浮动车数据采集技术和方法、GPS系统组成及定位原理、城市路网数据结构、道路网模型等。然后针对公交车实时采集的原始GPS数据存在的误差,分别进行了识别与处理,对缺失数据及异常数据进行修复或剔除；并且采用统计学的方法对大量的GPS数据进行直方图统计,筛选出高频轨迹点作为构建道路的核心点。在此基础上,提出了基于公交车GPS数据的单条道路路线生成方法,并对生成的道路路线通过去除异常点和拐角插值进行了进一步修正。在提取到单条道路路线后,本文提出了一种追加式路网生成方法,该方法通过不断追加道路路线来获取交叉口,切分路段,构建路网拓扑结构,从而逐渐形成完整的路网矢量地图。最后,本文利用济南市公交车辆的GPS数据进行了实验,实验结果表明该方法能够生成城市路网矢量地图,并且地图的位置精度较高,具有良好的应用价值。
[Abstract]:Urban road network vector map is the core component of ITS Intelligent Transportation system. It plays an important role in many application systems, such as location service, path planning, intelligent traffic guidance system and so on. However, with the rapid development of urban economy, the road network structure is becoming more and more complex and changing frequently, which brings more challenges to the maintenance of road network maps. At present, most of the road network vector maps rely on manual completion. Because of the high cost and long renewal period, it is difficult to reflect the changes of road network structure in time. Therefore, it is of great value to study a low cost and efficient method for automatic generation of urban road network map in recent years. More and more GPS vehicles are installed on buses and taxis. These floating cars are located on various roads in the city. They can upload the information about the vehicle's location in real time through GPS location. A large number of track data contain abundant road information, and these data have the advantages of high accuracy, wide road coverage, low acquisition cost and so on. This paper analyzes the characteristics of bus GPS data in depth. In this paper, a method of automatically generating urban road network map based on bus GPS data is presented. The data structure of urban road network, road network model and so on. Secondly, the errors of original GPS data collected by bus in real time are identified and processed, and the missing data and abnormal data are repaired or eliminated. A lot of GPS data are statistically analyzed by histogram, and the high-frequency track points are selected as the core points to construct the road. On this basis, a method of single road route generation based on bus GPS data is proposed. The generated road route is further modified by removing abnormal points and corner interpolation. After extracting a single road route, an additional road network generation method is proposed in this paper. In this method, the intersection is obtained by adding the road route continuously, the road sections are divided into sections, the topological structure of the road network is constructed, and the complete map of the road network vector is gradually formed. Finally, the GPS data of the public transport vehicles in Jinan are used to carry on the experiment in this paper. The experimental results show that this method can generate the vector map of urban road network, and the location accuracy of the map is high, and it has good application value.
【学位授予单位】：山东大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U495;TP311.13

【相似文献】