基于随机游走的道路选取研究

发布时间：2018-03-27 23:40

本文选题：道路网　切入点：随机游走　出处：《西南交通大学》2017年硕士论文

【摘要】：城市道路网是整个城市人流、物流运输的重要通道,是城市赖以生存与发展的重要物质基础。近年来,随着计算机与互联网技术的快速发展,城市道路网多尺度表达被广泛应用于电子地图导航与定位、地图制图学、空间数据挖掘与分析。如何对道路重要性进行评估和选取仍是其应用的核心问题之一。目前为止,道路的选取方法大多还是考虑到路网的空间形态特征和结构特征,对道路的交通运输功能关注较少;其中部分原因是大规模实时道路交通流数据不宜获得。随机游走作为研究信息或者能量扩散的基础模型,是研究一系列连续的随机步骤构成路径的统计模型,被广泛应用于多个学科。学者们基于随机游走算法构建了两种道路交通流模拟模型,即基于拓扑的交通流估算模型和基于几何的交通流估算模型。本论文依据人类行为学研究成果和道路网结构、形态,确定交通流估算模型参数,探究人类路径选择行为对道路交通流分布的影响,并对两种交通流估算模型进行优化。将估算的道路交通流与实际道路交通流做相关性分析,结果表明可以使用这两种交通流估算模型对道路交通流进行估算,基于几何的交通流估算模型在稳定性、准确性、要优于基于拓扑的交通流估算模型。道路交通流负载情况直接反映了道路的运输功能。那么,在制图综合中也应该考虑此功能。本文以香港岛道路网为例。分别以道路交通流、道路网长度、道路中介中心性、邻近中心性、集聚系数、度中心性作为选取依据,选取道路。并将选取结果与专家地图进行对比,结果表明交通流能同时顾及路网完整性和连通性,整体选取结果较好,可以作为道路选取指标之一。本论文的相关研究成果有助于理解道路交通流动态演化机制,为道路网制图综合和多尺度表达提供方法支持。
[Abstract]:The urban road network is the important passage of the whole city people flow, the logistics transportation, is the important material foundation of the city survival and development. In recent years, with the rapid development of computer and Internet technology, Multi-scale expression of urban road network is widely used in electronic map navigation and positioning, cartography, spatial data mining and analysis. How to evaluate and select the importance of road is still one of the core problems in its application. Most of the road selection methods take into account the spatial characteristics and structural characteristics of the road network, and pay less attention to the traffic and transportation function of the road. Part of the reason is that large-scale real-time road traffic flow data is not suitable for obtaining. As a basic model for studying information or energy diffusion, random walk is a statistical model for studying a series of continuous random steps to form a path. It has been widely used in many disciplines. Scholars have constructed two kinds of road traffic flow simulation models based on random walk algorithm. In this paper, the parameters of traffic flow estimation model are determined according to the results of human behavior research and the structure and shape of road network. This paper probes into the influence of human path selection behavior on the distribution of road traffic flow, and optimizes the two traffic flow estimation models. The correlation between the estimated road traffic flow and the actual road traffic flow is analyzed. The results show that the two traffic flow estimation models can be used to estimate the road traffic flow, and the geometric traffic flow estimation model is stable and accurate. It is better than the topology-based traffic flow estimation model. The traffic flow load directly reflects the transport function of the road. Therefore, this function should also be taken into account in cartographic generalization. This paper takes the Hong Kong Island Road Network as an example. Road network length, road intermediary centrality, proximity centrality, agglomeration coefficient, degree centrality are selected as the basis for selecting the road, and the selection results are compared with the expert map. The results show that the traffic flow can take into account the integrity and connectivity of the road network, and the overall selection result is good, which can be used as one of the road selection indicators. The related research results in this paper are helpful to understand the evolution mechanism of the road traffic flow state. It provides method support for road network cartography generalization and multi-scale representation.
【学位授予单位】：西南交通大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P208;P283

【相似文献】