【作者】 张辉；

【导师】 孙会君； 吴建军；

【作者基本信息】 北京交通大学 ， 系统分析与集成， 2012， 硕士

【摘要】 城市交通运输系统是一个复杂的巨系统,随着经济的发展、城市规模的扩大,城市交通变得越来越拥堵,不但给城市居民的出行带来了很大的困扰,也影响了城市的发展。复杂网络研究的蓬勃兴起为交通运输的研究提供了新的视角,交通网络作为一种复杂网络,其拓扑结构的好坏直接影响网络的承载能力。人们逐渐认识到,交通网络拓扑结构对网络交通流量分布有重要的影响,网络交通流量的分布依赖于底层网络拓扑。因此,研究不同网络拓扑下的流量分布规律对研究交通运输系统流量分布有重要意义。本文首先建立了不具有社区结构的相关性网络和具有社区结构的相关性网络,相关性网络包括正相关网络和负相关网络。对所构建的网络进行流量加载,然后按照用户平衡的方式达到平衡,对不具有社区结构的正、负相关性网络与不相关网络的流量分布和拥堵情况作比较,得出不相关网络的阻塞小于相关性网络；对于具有社区结构的网络也进行相同的比较,得出负相关网络的拥堵最小。研究说明网络的相关性对网络流量分布起着重要影响,这为我们规划、设计、改造交通网络提供指导。其次,本文对上述网络进行了级联失效研究。结果表明不具有社区结构的不相关网络能够更好的抵抗级联失效,而具有社区结构的负相关网络能够更好的抵御级联失效的影响。最后,本文提取了北京市的主要骨架路网,对路网进行了分析,结果表明北京市路网具有明显的负相关结构,具有14个社区。用户平衡配流的结果表明,一些主要路段随着流量的增加容易发生拥堵。这些路段是影响整个路网性能的关键路段,增加这些路段的承载能力可以有效的增加整个网络的性能。更多还原

【Abstract】 Urban transportation system is a complex giant system. With economic development, the expansion of the city, urban transport has become increasingly congested, which not only perplex urban residents to travel but also affect the development of city. The booming of the complex network research provides a new perspective for the study of transportation. As a complex network, urban road network topology directly impact on the carrying capacity of the network. People gradually realize that, the traffic network topology has important influence on the traffic flow distribution in the network and the traffic flow distribution depends on the underlying network topology. Therefore, studying the traffic flow distribution in different network topology is very meaningful.Firstly, we established two types of correlated network: with community structure and without community structure. Correlated network consists of the assortative network and disassortative network. Through the comparison between assortative, disassortative and uncorrelated networks with and without communities, we can obtain the flow distribution on the networks. We found that the uncorrelated network without community is better to resist congestion than correlated ones without community, while the disassortative network with community is better than other networks with community. The result can be used in city network evaluation, planning, transformation.Secondly, we study the cascading failure of the above mentioned network topologies. The studies of cascading failure on correlated networks show that the uncorrelated networks without community structure can be better to resist the cascading failure, however, the disassortative network with communities are better able to resist the cascading failure.Finally, we extract the main skeleton of Beijing road network, through calculating and analysis, the result shows that the Beijing road network is a disassortative network and has14communities. User equilibrium assignment results show that some of the major road sections are prone to be congested. These sections are keys to impact the whole network property and enhance the capacity of these sections can effectively enhance the capacity of the whole network.更多还原