基于复杂网络理论的交通流动态特性研究

发布时间：2018-05-02 05:51

本文选题：城市交通 + 路径选择算法　；参考：《西南交通大学》2014年博士论文

【摘要】：随着社会经济的快速发展,城市道路网络规模越来越庞大,但与此同时,汽车数量也在不断增长。从目前交通状况可以看出,城市交通系统的发展已无法满足当前人们日益增长的交通需求,进而引发的城市交通问题也愈发严重,交通拥堵、交通事故、道路阻塞等问题频频发生。从可持续发展的角度可知,城市规模不可能无限制地扩张下去,土地资源将越来越有限,越来越稀缺,如何在有限的道路资源条件下缓解城市交通拥堵并提高道路网络的交通承载能力已成为相关领域研究的重点和热点问题。为揭示城市交通流的内在机理,深入研究城市基础道路网络及建立在该网络之上的其他复杂系统(如公交系统)的拓扑结构特性,分析道路交通流及网络交通流的动态特性,有助于进一步探索缓解城市交通拥挤及提高城市路网吞吐量的交通诱导控制策略。所以,分析交通网络的拓扑结构复杂性及交通流复杂性对于城市交通问题研究至关重要。为此,本文围绕城市交通网络的结构特征及交通动态路由选择算法展开了深入研究。结合GIS与复杂网络理论,通过引入多粒度的概念系统地研究了城市交通网络的拓扑结构特性,在此基础上针对不同网络结构分析了交通流单向传输控制对整个网络交通过程的影响;深入分析了交通拥塞的产生机理,采用引力场理论实现了对交通流传输过程中节点之间相互作用的描述和定义,进而提出了基于节点引力场的动态路由选择算法。具体来讲,本论文的研究工作与成果主要有以下几个方面：1、结合GIS网络分析方法与复杂网络理论,对复杂交通网络模型的构建原理做了有益的探索并进行了相关统计分析,发现不同尺度下的路网均具有小世界和无标度特性,并通过引入多粒度的概念建立了城市道路多粒度复杂路网模型,进而分析了多粒度复杂路网的拓扑结构复杂性及整个城市道路网络的可靠性。研究发现,多粒度复杂路网模型具有无标度特性,有助于更为准确地分析城市道路网络的鲁棒性及脆弱性。2、为进一步分析城市交通系统的复杂性,针对城市公交系统,从乘客出行站点选择认知的角度出发,引入站点服务区的概念,并采用Voronoi图进行站点服务区的确定。在此基础上,分别针对公交站点网络和公交线路网络定义了公交服务可靠性指标及相应的攻击策略。试验证明,该公交系统可靠性分析方法可以较为准确地描述城市公交系统的鲁棒性及脆弱性。3、分析了实施交通流局部单向传递对整个网络交通状况的影响。分别以连接度和介数为约束条件,定义了两个交通流单向传递约束模型,并通过对ER随机网络、WS小世界网络和BA无标度网络等典型网络模型的交通模拟试验,得出了一个重要结论：对拥塞严重的节点实施交通流单向传递控制可以显著地提高ER随机网络和WS小世界网络的传输能力及缓解其网络拥塞程度,但不能有效地提高BA无标度网络的交通承载能力。该研究成果对实施城市交通(大部分城市交通网络被证实服从幂律分布,即为无标度网络)诱导控制提供了重要的决策参考依据。4、提出利用引力场理论来研究交通流传输过程中节点之间的相互作用,建立对交通引力场的描述,定义具有普适意义的节点引力场方程,定义了任意传输路径对数据包的引力计算公式,即将路径对数据包的引力表达为路径上所有节点对数据包的引力的平均值。在此基础上,提出了一种基于节点引力场的动态路由选择算法,即针对当前数据包的所有邻居节点到目标节点的最短路径,选择最短路径对数据包引力最大所对应的邻居节点作为下一个传输节点。模拟试验证明,该路径选择算法较大地提高了整个网络的传输能力,显著地缓解了网络的拥塞程度。5、为深入探讨基于引力场理论的路由选择策略的交通流动力学特性,引入路径感知深度的概念,定义了在路径感知深度约束下传输路径对数据包的引力计算公式,并给出了相应的引力路由选择算法。试验结果揭示了一个重要的动力学现象：当路径感知深度大于网络平均距离长度时,该路由选择算法可以显著地提高整个网络的传输能力,且网络传输性能将不再随路径感知深度的持续增大而变化,网络传输性能将进入稳定状态。6、从引力均衡的角度,引入标准差的基本思想,认为在节点引力场作用下或许存在一个最佳的临界引力,且在该引力下的路由选择过程更为高效。基于这种假设,建立了一个反映节点引力离散程度的数学模型,且基于该数学模型提出了一种新的引力场路由选择算法。试验结果表明,该路由选择算法显著地提高了网络的传输能力,有效地均衡了网络交通负载,在一定程度上该算法的性能优于上述基于路径节点引力平均值的路由选择算法。
[Abstract]:With the rapid development of social economy, the scale of urban road network is becoming larger and larger, but at the same time, the number of cars is increasing. From the current traffic conditions, it can be seen that the development of urban traffic system can not meet the increasing demand for traffic, and the urban traffic problems are becoming more and more serious, traffic congestion, Traffic accidents, road congestion and other problems occur frequently. From the perspective of sustainable development, the scale of the city can not be expanded unrestricted, the land resources will become more and more limited and more and more scarce. How to alleviate traffic congestion under the limited road resources and raise the traffic carrying capacity of road network has become a relevant leader. In order to reveal the inner mechanism of urban traffic flow and to study the topological structure characteristics of urban basic road network and other complex systems (such as bus system) on the network, the dynamic characteristics of road traffic flow and network traffic flow are analyzed, which will help to further explore the relief of urban traffic. Therefore, the analysis of topology complexity and traffic flow complexity of traffic network is very important to the research of urban traffic problems. Therefore, this paper studies the structure characteristics of urban traffic network and the algorithm of traffic dynamic routing selection. It combines GIS and complexity. In the network theory, the topological structure of urban traffic network is studied by introducing the concept of multi granularity. On this basis, the influence of traffic flow control on the whole network traffic process is analyzed according to different network structures. The mechanism of traffic congestion is analyzed, and the traffic flow is realized by the gravitational field theory. The interaction between nodes in the transmission process is described and defined, and then the dynamic routing algorithm based on the node gravitational field is proposed. In particular, the research work and results of this paper are mainly as follows: 1, combining the GIS network analysis method and the complex network theory, the construction principle of the complex traffic network model is made. It is found that the road network under different scales has the characteristics of small world and scale-free, and the multi granularity complex road network model of urban road is established by introducing the concept of multi granularity, and then the complexity of topology structure of multi granularity complex road network and the reliability of the whole urban road network are analyzed. It is found that the multi granularity and complex road network model has the characteristics of scale free, which can help to analyze the robustness and vulnerability of urban road network more accurately. In order to further analyze the complexity of urban traffic system, the concept of site service area is introduced to the urban bus system, the concept of site service area is introduced, and the.2 is adopted, and Vo is adopted. On the basis of this, the reliability index of bus service and the corresponding attack strategy are defined on the basis of the bus station network and the bus line network. The test proves that the reliability analysis method of the bus system can describe the robustness and vulnerability of the urban public transportation system more accurately, and the analysis of.3 is more accurate. The effect of traffic flow local one-way transmission on the whole network traffic condition is implemented. Two traffic flow one-way transmission constraint models are defined respectively with the connection degree and the number of medials as constraints, and an important conclusion is drawn through the traffic simulation experiments of typical network models such as ER random network, WS small world network and BA scale-free network. The implementation of traffic flow control for congested nodes can significantly improve the transmission capacity of the ER random network and WS small world network and alleviate the network congestion, but it can not effectively improve the traffic carrying capacity of the BA scale-free network. From the power law distribution, which is an important reference basis for decision making for the scale-free network,.4, the gravitational field theory is used to study the interaction between nodes in the transport process of traffic flow, to establish a description of the gravitational field of traffic, to define the universal gravitational field equation with a universal meaning, and to define the arbitrary transmission path to the data packet. The gravitational calculation formula in which the gravitational expression of the path to the packet is expressed as the average value of the gravitational force of all nodes on the path on the path. On this basis, a dynamic routing algorithm based on the node gravitational field is proposed, that is to select the shortest path logarithm for the shortest path of all the neighbor nodes of the current packet to the target node. The simulation experiment shows that the path selection algorithm greatly improves the transmission capacity of the whole network, significantly alleviates the network congestion.5, and discusses the traffic flow mechanics characteristics of the routing strategy based on gravitational field theory, and introduces the path sense. In the concept of depth of knowledge, the gravitational calculation formula of the transmission path to data packets is defined under the path perception depth constraint, and the corresponding gravitational routing algorithm is given. The results of the experiment reveal an important dynamic phenomenon: the routing algorithm can be significantly raised when the path perception depth is greater than the average distance of the network. The transmission capability of the whole network and the transmission performance of the network will no longer vary with the continuous increase of path perception depth. The network transmission performance will enter the stable state.6. From the angle of gravitational equilibrium, the basic idea of the standard deviation is introduced. It is considered that there may be a best critical gravity under the action of the gravitational field of the node and under the gravitational force. The route selection process is more efficient. Based on this hypothesis, a mathematical model is established to reflect the degree of gravity discretization of nodes, and a new routing algorithm for gravitational field is proposed based on this model. The experimental results show that the routing algorithm significantly improves the transmission capacity of the network and effectively balances the network traffic negative. To a certain extent, the performance of the algorithm is better than that of the routing average algorithm based on the path node.

【学位授予单位】：西南交通大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：U491.112

【参考文献】