基于关键交叉口的城市道路交通小区划分方法研究

发布时间：2018-03-25 08:48

本文选题：凝聚算法　切入点：关键节点　出处：《吉林大学》2017年硕士论文

【摘要】：城市道路交通网络是一个离散、强耦合、非线性、并具有随机特性的复杂网络,通过交通小区划分可以将其解析为若干个区域,并针对各个小区实施相应的交通管理和诱导策略,从宏观角度对交通区域进行协调优化。因此,交通小区划分在降低系统复杂性方面具有重要的意义。为了便于分区缓解交通拥堵、实施区域交通的同步控制和接力诱导,需要识别出以关键交叉口为核心的交通小区。考虑到城市道路交通网络的复杂性,论文利用复杂网络社区挖掘理论来研究基于关键交叉口的城市道路交通小区划分方法,主要研究内容如下:首先,介绍了与论文研究内容相关的复杂网络特征度量参数和网络特性,对复杂网络社团挖掘方法进行了对比分析,特别是凝聚算法。论文介绍了CNM凝聚算法、基于Laplace谱分析的凝聚算法和基于局部模块度的凝聚算法的思路和实施步骤,并选取长春市牡丹街、桂林路、丰顺街、清华街包围的路网为实例,从模块度和时间复杂度的角度对三种方法进行比较,以便确定适合于城市道路交通网络的交通小区划分方法。其次,论文介绍了根据道路交通特征参数和网络拓扑结构特征识别关键交叉口的方法,并研究基于关键交叉口的城市道路交通网络交通小区划分方法。论文以节点间的关联度为道路交通网络边权值,以局部模块度最大化为准则,提出了基于关键节点的交通划分方法,并以沈阳市区域路网进行例对该方法进行验证。最后,论文研究城市道路交通网络中重叠区域的处理方法。针对网络中不同类型的重叠区域,分为两个区域重叠的处理方法和多个区域重叠的处理方法。两个区域的重叠的处理方法以相似度指标为适应度函数,并构造基于遗传算法的求解方法。多个区域重叠的处理方法以重叠区域模块度为指标进行预处理,将多重叠变为两重叠。以第三章获得的带有重叠区域的小区划分结果,进行重叠区域处理的实例验证,结果表明该方法可靠。
[Abstract]:Urban road traffic network is a discrete, strongly coupled, nonlinear and stochastic complex network. And implement the corresponding traffic management and guidance strategy for each district, and optimize the traffic area from the macro point of view. In order to reduce traffic congestion, traffic district division is of great significance in reducing the complexity of the system. In order to ease traffic congestion, regional traffic synchronization control and relay guidance are implemented. Considering the complexity of urban road traffic network, this paper uses the complex network community mining theory to study the method of urban road traffic district division based on key intersection. The main research contents are as follows: firstly, the measurement parameters of complex network features and network characteristics related to the contents of this paper are introduced, and the methods of community mining in complex networks are compared and analyzed. Especially the agglomeration algorithm. This paper introduces the CNM condensation algorithm, the condensing algorithm based on Laplace spectrum analysis and the condensing algorithm based on local modularity, and selects the Mudan Street, Guilin Road, Fengshun Street of Changchun City. The road network surrounded by Qinghua Street is taken as an example to compare the three methods from the angle of modularity and time complexity in order to determine the traffic district division method suitable for urban road traffic network. Secondly, This paper introduces a method to identify key intersections according to road traffic characteristic parameters and network topology features. The paper takes the correlation degree of the nodes as the boundary weight of the road traffic network, and takes the maximization of the local modularity as the criterion. A traffic partition method based on key nodes is proposed, and the method is verified by an example of Shenyang regional road network. This paper studies the methods of dealing with overlapping areas in urban road traffic networks, aiming at different types of overlapping areas in urban traffic networks. It is divided into two regions overlapping processing methods and multiple regions overlapping processing methods. The two regions overlap processing methods take similarity index as fitness function. A genetic algorithm based method is constructed. The processing method of multiple overlapping regions is preprocessed with the module degree of overlapped regions to change multiple overlaps into two overlaps. The results of cell division with overlapped regions obtained in Chapter 3 are used. The results show that the method is reliable.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.12

【相似文献】