基于传输贡献矩阵的城市路网节点重要性评估方法  

Evaluations of node importance of urban road network based on transmission contribution matrix

作　　者：陈亮 王京京 周志华 李巧茹 陈明明 CHEN Liang;WANG Jingjing;ZHOU Zhihua;LI Qiaoru;CHEN Mingming(School of Civil Engineering and Transportation, Hebei University of Technology, Tianjin 300401, China;Guangzhou Transport Planning Research Institute, Guangzhou 510230, China)

机构地区：[1]河北工业大学土木与交通学院,天津300401 [2]广州市交通规划研究院,广州510230

出　　处：《科技导报》2018年第6期105-111,共7页Science & Technology Review

基　　金：国家自然科学基金项目(51678212);北京城市交通协同创新中心项目

摘　　要：在城市道路中,通常采用交叉口流量或道路连接数等单一指标确定路口重要性,很难保证突发事件下道路网络运行的可靠性。鉴于城市道路网络的无标度特性,以交通阻抗为权重,将城市道路网络抽象为加权网络,提出节点重要度传输贡献矩阵方法,评估道路交叉口重要性。该方法综合考虑交叉口所处路网位置、道路连接数及关联路段交通阻抗,采用节点效率值表示其位置信息,通过融合节点度值和交通阻抗值,构建交通信息传播过程中节点之间的重要度关联,以表示相邻路口间重要性依赖关系。以天津市区域路网为例计算各交叉口重要度,采取连续移除重要节点方式,模拟连锁故障过程,并将结果与节点收缩法进行对比。实证分析表明：在移除最重要的8个节点后,路网最大连通子图规模较节点收缩法降低9.10%,验证了该方法的有效性。Some important intersection nodes in the urban road network are generally defined by a single indicator such as the traffic flow or the number of connected links. It is difficult to ensure the reliability of the road network once the traffic incident in the hub node happens. In this paper, according to the free scale features, the urban road network is simplified as a weighted complex network, in which the traffic impedance is taken as the weight parameter, and a method of the weighted network node importance transmission contribution matrix is proposed. The method comprehensively combines the intersection location in the road network, the number of connected links, and the traffic impedance of associated links, to evaluate the intersection node importance. The node efficiency value is used to represent the information related with the intersection location, meanwhile, the node degree value and the traffic impedance value are fused to characterize the importance dependencies of adjacent intersections in the process of the traffic information transmission. A regional road network in Tianjin is used as a test sample to validate the proposed approach. The cascading failure is simulated on the network by continuously removing important nodes. Simulation results show that the largest subgraph size based on the proposed method falls by 9.10%, as compared with the weighted node contraction method when the top 8 important nodes are continuously removed. These results provide a theoretical basis for the congestion warning for the urban traffic network.

关 键 词：城市路网 交叉口节点 交通阻抗 复杂网络 

分 类 号：O157.5[理学—数学] U491[理学—基础数学]