基于元胞自动机的交通流模型研究
发布时间：2018-11-01 11:00
【摘要】： 随着国民经济的发展，人们对各类交通的需求不断增加。而交通流量的增长与现有道路状况之间的矛盾日益突出，成为限制社会经济发展的主要因素之一。交通流理论的研究目标是要建立能够描述实际交通一般特性的交通流模型，揭示交通流的基本规律。开展交通流理论的研究，不仅有着深远的科学意义，而且具有重要的工程应用价值。 论文在元胞自动机理论的基础上，对交通流模型的建模、仿真及其应用等方面进行了研究，主要包括以下内容： 1、对元胞自动机交通流模型进行了分类、仿真以及分析评价。根据元胞自动机交通流模型中车辆占据的元胞个数、车道数量以及模型维数将其分为单细胞和多细胞TCA模型、单车道和多车道TCA模型、一维和二维TCA模型。在此分类的基础上，，应用Java程序对每种TCA模型的时空关系和交通流基本图进行了细致的仿真，并结合仿真结果对模型进行了分析和评价。 2、建立了一种改进的NS模型，并对模型的随机特性进行了研究和应用。将一组恒速控制条件引入NS模型的速度控制规则之中，提出了一组新的速度控制规则。在此基础上，将随机减速概率p从新模型中提取出来，作为一个随机参数植入一阶连续介质理论交通流模型(LWR模型)之中，并重构了其交通流基本图。数值模拟结果和实测数据对比表明，这种方法能够较为真实的再现实际交通流特征，并且很好的诠释了交通流的流体力学特性。 3、将车辆的OD属性引入二维的BML模型中，根据系统内车辆是否守恒建立了两种二维TCA模型。定义了模型中车辆的起讫点、起讫距离以及起讫距离的测量方法，并给出了模型中起讫距离概率依从的三种分布方式，即指数分布、幂分布和均匀分布。在此基础上给出了模型中车辆的基本运动规则，建立了车辆守恒的扩展BML模型和非守恒的扩展BML模型。对新建模型以及模型中参数之间关系进行了数值模拟，结果表明：通过合理的选择起讫距离的概率分布形式以及调节模型中参数的大小，可以提高城市道路的通行能力，减小车辆的出行时间。 4、基于元胞自动机的高效运算能力，结合出发时间选择模型、动态路线选择模型和动态网络装载模型，借鉴动态建模的思想，构建了基于元胞自动机的动态交通分配系统框架。 5、在GIS技术支持下，根据以上构建的动态交通分配系统框架，建立了一种用于动态交通分配的多模式、可扩展的路网数据结构。将此数据结构转化为GIS中的属性数据，以实际路网为背景进行了动态交通分配，并根据分配结果对当前路网做出了分析、评价和改进建议。 最后，对论文的工作进行了总结，并指出了下一步需要继续研究的内容。
[Abstract]:With the development of national economy, people's demand for all kinds of traffic is increasing. The contradiction between the growth of traffic flow and the existing road condition is becoming more and more prominent, which has become one of the main factors restricting the development of social economy. The research goal of traffic flow theory is to establish a traffic flow model which can describe the general characteristics of real traffic and to reveal the basic laws of traffic flow. The research of traffic flow theory not only has profound scientific significance, but also has important engineering application value. Based on the theory of cellular automata, this paper studies the modeling, simulation and application of traffic flow model. The main contents are as follows: 1. The traffic flow model of cellular automata is classified. Simulation and analysis and evaluation. According to the number of cells occupied by vehicles, the number of lanes and the dimension of the model, the model can be divided into single cell and multi-cell TCA model, single-lane and multi-lane TCA model, one-dimensional and two-dimensional TCA model. On the basis of this classification, the spatio-temporal relationship of each TCA model and the basic traffic flow diagram are simulated with Java program, and the model is analyzed and evaluated with the simulation results. 2. An improved NS model is established, and the stochastic characteristics of the model are studied and applied. A set of constant speed control conditions is introduced into the speed control rules of NS model, and a new set of speed control rules is proposed. On this basis, the random deceleration probability p is extracted from the new model as a random parameter and implanted into the first-order continuum theoretical traffic flow model (LWR model), and its basic traffic flow graph is reconstructed. The comparison between the numerical simulation results and the measured data shows that this method can reproduce the real traffic flow characteristics and explain the hydrodynamic characteristics of the traffic flow very well. 3. The OD attribute of the vehicle is introduced into the two-dimensional BML model, and two two-dimensional TCA models are established according to whether the vehicle in the system is conserved or not. In this paper, the methods of measuring the starting point, starting distance and starting distance of vehicle in the model are defined, and three distribution modes of probability compliance of starting and end distance in the model are given, that is, exponential distribution, power distribution and uniform distribution. On this basis, the basic motion rules of the vehicle in the model are given, and the extended BML model of vehicle conservation and the extended BML model of non-conserved vehicle are established. The new model and the relationship between the parameters in the model are numerically simulated. The results show that the traffic capacity of urban roads can be improved by selecting the probability distribution form of the starting distance and adjusting the size of the parameters in the model. Reduce vehicle travel time. 4. Based on the efficient computing ability of cellular automata, combined with the departure time selection model, the dynamic route selection model and the dynamic network loading model, the idea of dynamic modeling is used for reference. The framework of dynamic traffic assignment system based on cellular automata is constructed. 5. With the support of GIS technology, according to the above framework of dynamic traffic assignment system, a multi-mode and extensible road network data structure for dynamic traffic assignment is established. The data structure is transformed into attribute data in GIS, and dynamic traffic assignment is carried out under the background of the actual road network. According to the result of the distribution, the analysis, evaluation and improvement suggestions are made for the current road network. Finally, the paper summarizes the work, and points out the next step to continue to study the content.
【学位授予单位】：山东科技大学
【学位级别】：博士
【学位授予年份】：2007
【分类号】：U491
本文编号：2303716
[Abstract]:With the development of national economy, people's demand for all kinds of traffic is increasing. The contradiction between the growth of traffic flow and the existing road condition is becoming more and more prominent, which has become one of the main factors restricting the development of social economy. The research goal of traffic flow theory is to establish a traffic flow model which can describe the general characteristics of real traffic and to reveal the basic laws of traffic flow. The research of traffic flow theory not only has profound scientific significance, but also has important engineering application value. Based on the theory of cellular automata, this paper studies the modeling, simulation and application of traffic flow model. The main contents are as follows: 1. The traffic flow model of cellular automata is classified. Simulation and analysis and evaluation. According to the number of cells occupied by vehicles, the number of lanes and the dimension of the model, the model can be divided into single cell and multi-cell TCA model, single-lane and multi-lane TCA model, one-dimensional and two-dimensional TCA model. On the basis of this classification, the spatio-temporal relationship of each TCA model and the basic traffic flow diagram are simulated with Java program, and the model is analyzed and evaluated with the simulation results. 2. An improved NS model is established, and the stochastic characteristics of the model are studied and applied. A set of constant speed control conditions is introduced into the speed control rules of NS model, and a new set of speed control rules is proposed. On this basis, the random deceleration probability p is extracted from the new model as a random parameter and implanted into the first-order continuum theoretical traffic flow model (LWR model), and its basic traffic flow graph is reconstructed. The comparison between the numerical simulation results and the measured data shows that this method can reproduce the real traffic flow characteristics and explain the hydrodynamic characteristics of the traffic flow very well. 3. The OD attribute of the vehicle is introduced into the two-dimensional BML model, and two two-dimensional TCA models are established according to whether the vehicle in the system is conserved or not. In this paper, the methods of measuring the starting point, starting distance and starting distance of vehicle in the model are defined, and three distribution modes of probability compliance of starting and end distance in the model are given, that is, exponential distribution, power distribution and uniform distribution. On this basis, the basic motion rules of the vehicle in the model are given, and the extended BML model of vehicle conservation and the extended BML model of non-conserved vehicle are established. The new model and the relationship between the parameters in the model are numerically simulated. The results show that the traffic capacity of urban roads can be improved by selecting the probability distribution form of the starting distance and adjusting the size of the parameters in the model. Reduce vehicle travel time. 4. Based on the efficient computing ability of cellular automata, combined with the departure time selection model, the dynamic route selection model and the dynamic network loading model, the idea of dynamic modeling is used for reference. The framework of dynamic traffic assignment system based on cellular automata is constructed. 5. With the support of GIS technology, according to the above framework of dynamic traffic assignment system, a multi-mode and extensible road network data structure for dynamic traffic assignment is established. The data structure is transformed into attribute data in GIS, and dynamic traffic assignment is carried out under the background of the actual road network. According to the result of the distribution, the analysis, evaluation and improvement suggestions are made for the current road network. Finally, the paper summarizes the work, and points out the next step to continue to study the content.
【学位授予单位】：山东科技大学
【学位级别】：博士
【学位授予年份】：2007
【分类号】：U491
【引证文献】
相关期刊论文 前1条
1 陈永;王晓明;党建武;;基于元胞自动机理论的快速公交系统能耗研究[J];计算机应用;2012年03期
相关博士学位论文 前3条
1 陈晨;基于细胞自动机方法的城市交通流模拟及优化[D];华中科技大学;2010年
2 柯姜岑;基于元胞自动机的水运枢纽运输组织研究[D];武汉理工大学;2012年
3 袁绍欣;城市交通拥堵传播机理及其控制策略研究[D];长安大学;2012年
相关硕士学位论文 前2条
1 吴国辉;北京站交通衔接仿真与优化研究[D];北京交通大学;2011年
2 宋升;基于元胞自动机的集装箱码头前沿路网结构建模与分析[D];武汉理工大学;2011年
本文编号：2303716
本文链接：https://www.wllwen.com/kejilunwen/jiaotonggongchenglunwen/2303716.html
