基于出租车轨迹的路网交通流建模研究

发布时间：2018-03-26 15:00

本文选题：交通流网络LWR模型　切入点：GPS　出处：《北京交通大学》2017年硕士论文

【摘要】：城市交通一直是城市经济发展和人们日常活动的关键所在,随着经济发展的逐步加快,交通拥堵现象日趋严重,交通基础设施建设已经不能满足人们的出行需求。同时城市交通状况愈发的复杂化,因此人们希望找到一种方法从整体上反映城市交通的变化规律。宏观交通流模型由于其反映了整个交通网络状况,对总体把控交通状态具有指导性意义,所以在交通流模型中显得尤其重要。在宏观交通流模型中,LWR模型具有参数少、易于计算仿真等优点,是目前最为广泛使用的方法。但一般情况下,LWR模型适用于结构相对简单的交通仿真,例如连续二维平面、直线道路等等。而实际上城市路网结构复杂,干扰因素多,把LWR模型与真实路网结合起来一直存在诸多难点。所以本文基于LWR模型对于实际道路的仿真在宏观交通流研究中很有意义。本文基于交通流网络LWR模型,阐述了模型结构、对应数值格式以及Riemann等相关问题。同时结合出租车GPS数据,获取相关交通信息。结合实际路网数据,完成了路网数据提取工作,并在此基础上进行出租车轨迹数据匹配,获得路网相关交通流参数。最终将交通流网络LWR模型、GPS数据以及路网数据三者进行融合,得到了基于路网交通流LWR模型的仿真结果。本文主要工作内容包括:1.介绍了宏观LWR模型的发展历程,讨论了交通流模型的一般形式,阐述了多种经典模型。然后从计算的角度出发,介绍了模型的基本性质以及数值格式。并在一阶LWR模型的基础上说明了交通流网络LWR模型的构造方法、Riemann问题以及模型的计算方法。2.从数据的角度出发,以出租车GPS数据为基础,进行了数据预处理、数据挖掘,特征提取等工作;以路网数据为基础,完成了路网数据预处理、道路方向确定、路网细化等工作。最后将二者结合,完成了道路匹配。3.基于交通流网络LWR模型在路网上进行仿真,并根据交通流参数对模型内部参数进行了调整。本文的主要成果包括:1.基于北京市道路信息以及特征提取结果得到了北京市主要城市道路网。2.通过匹配获取了路网中各个路段出租车速度、流量等交通流参数。3.完成了交通流网络LWR模型在路网上进行仿真。文章中共包含图25幅,表10个,参考文献60篇。包含了数据处理、路网提取、模型仿真等各部分内容。文章的数据处理过程对于交通流模型在实际路网上仿真提供了可靠依据,得到的仿真结果对于LWR模型在实际城市交通网络中的应用以及宏观交通流研究都具有一定的意义。
[Abstract]:Urban traffic has always been the key to the development of urban economy and people's daily activities. With the gradual acceleration of economic development, traffic congestion is becoming more and more serious. Transportation infrastructure has not been able to meet people's travel needs. Meanwhile, the urban transportation situation is becoming more and more complicated. Therefore, people hope to find a way to reflect the changing law of urban traffic on the whole. Because the macroscopic traffic flow model reflects the whole traffic network condition, it is of guiding significance to control the traffic state as a whole. So it is especially important in the traffic flow model. In the macroscopic traffic flow model, the LWR model has the advantages of less parameters, easy calculation and simulation, etc. It is the most widely used method at present. But in general, the LWR model is suitable for traffic simulation with relatively simple structure, such as continuous two-dimensional plane, straight road, etc. In fact, the structure of urban road network is complex, and there are many interference factors. There are many difficulties in combining the LWR model with the real road network. Therefore, the simulation based on the LWR model is of great significance in the study of the macroscopic traffic flow. Based on the LWR model of the traffic flow network, this paper expounds the structure of the model. Corresponding numerical format and Riemann and other related problems. At the same time combined with taxi GPS data to obtain relevant traffic information. Combined with the actual road network data, completed the road network data extraction work, and on the basis of the taxi track data matching, Finally, the traffic flow network LWR model and road network data are fused together. The simulation results of traffic flow LWR model based on road network are obtained. The main work of this paper includes: 1. The development history of macro LWR model is introduced, and the general form of traffic flow model is discussed. Several classical models are described. Then, from the point of view of calculation, This paper introduces the basic properties and numerical format of the model, and on the basis of the first-order LWR model, explains the construction method of the LWR model of traffic flow network and the calculation method of the model. 2. From the point of view of the data, based on the data of the taxi GPS, the paper introduces the method of constructing the LWR model of the traffic flow network and the calculation method of the model. Data preprocessing, data mining, feature extraction and so on are carried out. Based on road network data, road network data preprocessing, road direction determination, road network refinement and so on are completed. The road matching. 3. Based on the traffic flow network LWR model, the road network simulation is carried out. The main results of this paper are as follows: 1. Based on the road information and feature extraction of Beijing, the road network of main cities in Beijing is obtained. 2. The road network is obtained by matching. Taxi speed in various sections of the network, 3. The traffic flow network LWR model is simulated on the road network. The paper includes 25 figures, 10 tables, 60 references, including data processing, road network extraction, and so on. The data processing process of the paper provides a reliable basis for the traffic flow model simulation on the actual road network. The simulation results are significant for the application of LWR model in urban traffic network and the study of macroscopic traffic flow.
【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】