基于手机基站数据的城市路网使用需求分析

发布时间：2018-08-24 07:31

【摘要】：我国经济稳步发展的同时,交通拥堵成为一个困扰各大大小小城市的问题。单纯地拓展交通线路并不是一个可持续发展的办法,只有把握整个路网的使用情况才能制定合理的应对措施。然而由于覆盖的样本有限,基于现代科技的交通流量检测方法(波频检测技术、磁频检测技术和视频检测技术)和基于GPS的交通流量检测方法均不能较全面地反映道路的使用情况。另一方面,有着极高持有率的手机为交通信息的获取提供了大量的数据。因此在深入了解现有道路使用需求计算方法的基础上,本文使用手机基站数据计算和分析南京城市路网的使用需求。其主要工作包括：1) 对基站进行层次聚类以缓解过去基站覆盖范围模型存在的问题；并使用SQL语句简洁地实现了OD矩阵的计算。2)依据开源地图OSM完成了南京城市路网的构建。3)针对使用Dijkstra最短路径算法模拟车辆出行线路所存在的不足,设计了基于转移概率的路径选择算法。4) 为配合本文的路径选择算法,设计了新的区域到路网的匹配方法；并在此基础上计算了各道路的使用需求及其来源。在缺乏真实道路使用需求数据的情况下,本文根据日常生活经验来验证计算得到的结果。主要的发现有：1) OD区域按其人数变化趋势可以分成两类,且晚上人多的一类大致属于居民区。2) 计算得到的OD量与道路使用需求量的变化趋势符合人们的通勤规律。具体表现为,工作日出行早高峰在7：30到9：00之间；而晚高峰在17：00到19：00之间；且在中午12：00左右会出现短距离出行的小高峰。3) 道路使用需求量大小与其来源数量呈正相关,且源头对道路使用需求的贡献大致服从二八定律。4) 计算得到的高使用需求道路对现实中的拥堵道路有着相当高的召回率,如南京长江大桥、中央路、玄武湖隧道和应天大街高架等等。上述的对比和发现证明了本文设计的计算方法的有效性。最后,结合道路使用需求来源的计算,本文给出了其在交通领域的潜在应用,包括潜在拥堵识别、拥堵成因分析和城市交通规划辅助决策支持。
[Abstract]:With the steady development of economy in our country, traffic congestion has become a problem that puzzles cities large and small. Simply expanding traffic lines is not a sustainable development method. Only by grasping the use of the whole road network can we formulate reasonable countermeasures. However, due to the limited number of samples covered, a new method of traffic flow detection based on modern science and technology (wave frequency detection, Neither the magnetic frequency detection technology nor the video detection technique] and the traffic flow detection method based on GPS can not fully reflect the road usage. On the other hand, mobile phones with extremely high holding rate provide a lot of data for obtaining traffic information. Therefore, based on the deep understanding of the existing road demand calculation methods, this paper uses mobile phone base station data to calculate and analyze the demand of Nanjing urban road network. The main work includes: (1) hierarchical clustering of base stations to alleviate the problems existing in the past base station coverage model; The OD matrix is calculated by using SQL sentence. 2) based on the open source map OSM, the construction of Nanjing urban road network is completed. 3) aiming at the shortcomings of using Dijkstra shortest path algorithm to simulate vehicle travel routes. A path selection algorithm based on transition probability is designed. In order to match the path selection algorithm in this paper, a new matching method of region to road network is designed, and on this basis, the demand for each road and its source are calculated. In the absence of real road use demand data, this paper verifies the calculated results according to daily life experience. The main findings are as follows: (1) the OD region can be divided into two categories according to its population change trend, and the type with large number of people at night belongs to the residential area approximately. 2) the change trend of OD quantity and road use demand obtained by calculation accords with the commuting law of people. It is shown that the peak of working day travel is between 7:30 and 9:00, while the evening peak is between 17:00 and 9:00. And there will be a small peak of short distance travel around 12:00) the demand for road use is positively correlated with the number of sources. And the contribution of the source to the demand for road use is generally satisfied with the high demand for roads calculated from law .4). The roads with high demand have a high recall rate to the congested roads in reality, such as the Nanjing Yangtze River Bridge, the Central Road, Xuanwu Lake Tunnel and Yingtian Street elevated and so on. The comparison and discovery above prove the validity of the calculation method designed in this paper. Finally, combined with the calculation of the demand sources of road use, the potential applications in the field of traffic, including potential congestion identification, congestion cause analysis and urban traffic planning decision support, are given in this paper.
【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

【相似文献】