【作者】 陈峰；

【导师】 贾元华；

【作者基本信息】 北京交通大学 ， 系统工程， 2012， 博士

【摘要】 高速公路与关联城市快速路结合部路网拥堵日益严重,而相应的交通管理和控制却由于管理体制分割存在明显“盲区”。在大城市结合部路网内,高速公路联络线和城市快速路负荷集中度较高,不同性质的车流快速交汇、转换和大流量对结合部路网的冲击更易造成交通阻塞并快速蔓延,局部路段拥堵加剧路网整体通行能力及可靠性降低。应对大城市结合部路网交通拥堵已成为亟需解决的现实问题,同时也是城市交通规划与管理学科面临的一个复杂系统问题。为此,本文应用系统工程的思想,综合运用人工智能、交通工程、数理统计等交叉学科的理论、技术和方法,在仿真技术的支持下,针对结合部路网交通控制问题及其结构特征,运用智能协同控制技术和集成控制技术来处理这一复杂的系统问题。本论文的主要研究成果包括：(1)充分考虑高速公路与城市快速路两个不同技术等级标准、不同管理主体以及交通数据异构等特点,对本文提出的异构数据融合模式进行探讨,然后利用灰色关联分析方法对路网内交通流、断面和匝道间的关联度变化规律展开分析,并通过对趋势关联度进行一阶差分修正和二阶差分修正,建立了结合部路网二阶趋势关联度模型。(2)结合高速公路进出城路段交通流的不确定性、动态性、关联性和互补性等特点,针对多断面交通数据,构建基于径向基神经网络的短时交通流预测模型；依据交通流连续性假设,建立了交通压力公式,构建基于交通流量短时预测结果的结合部路网多断面动态服务水平判定模型。(3)以预防拥堵、推迟失效和维持结合部路网通行能力为优化目标,提出结合部路网分层递阶智能控制结构；然后采用SWARM的控制思想建立基于分层递阶智能控制结构的全系统自适应多匝道智能协同控制模型。(4)以路况、气象条件等为交通运行安全和效率的约束条件,将车辆行驶速度作为主要控制变量,从交通流状态函数关系入手,建立基于支持向量机回归的主线限速控制模型,并设计结合部路网主线限速控制器。(5)考虑结合部路网内主线和匝道对交通控制的影响,运用最优控制理论和多匝道智能协同控制理论及建模方法,建立结合部路网主线关联多匝道集成控制模型,并通过实证分析,证明结合部路网考虑主线和多匝道集成控制能够有效地消除或缓解交通拥挤和维持主线车流稳定,提高结合部路网的整体通行能力的优化目标及其效果。更多还原

【Abstract】 Traffic congestion in metropolitan periphery road networks are growing seriously. However, these areas are lack of control due to the separation of the management system. In the junction road network, the expressway and urban roads are highly loaded and concentrated. Furthermore, the traffic volumes with different properties interchange and conversion with each other, which easily leads to traffic congestion that will quickly spread. The congestion in these parts could reduce the overall capacity of the road network, which has been obviously limited the efficiency of the city’s operations. Therefore, the issue of the road junction has become an important factor which restricts the development of city and needs to be solved urgently. Meanwhile, it is a complex systematic problem faced in the traffic planning and management. This paper utilizes the theory of system engineering and combines it with artificial intelligence, traffic engineering, mathematical statistics. According to the structural features of the traffic control on the junction road network, it attempts to solve this complex systematic problem with the intelligent cooperative control technology and integrated control technology based on the emulation technique.The research findings of this paper include:(1) The different standards, managements and isomerism of the traffic data of expressway and urban road have been full considered, on the basis of which the blending mode of isomerous data was discussed. After that, this paper analyzed the changing regularity of relevance among traffic volume, sections and ramps with grey correlation analysis. It also made the first-order difference modification and second-order difference modification on trend relational degree and built the second-order trend relational model of the junction road network.(2) Associating with the uncertainty, dynamism, relevancy and complementarities of the traffic volumes that enter and exit the city from the expressway and according to the traffic data of sections, the short-term traffic volume prediction model was built sections traffic data based on Radial Basis Function Neural Network. Then, on the basis of the assumption that the traffic volume is continuous, the traffic pressure formula was built, so was the decision model of multi-section dynamic service levels of junction road network based on the results of the short-term traffic volume prediction.(3) This paper has proposed the hierarchical intelligent control structure of junction road network according to the optimization objectives, which are preventing the congestion, postponing the invalidation of traffic control and improving the traffic capacity of the junction road network. Then this paper built the self-adapted and full-systematic intelligent cooperative control model based on hierarchical intelligent control model with the applying of SWARM.(4) In this paper, road conditions and climate are taken as constraint conditions of traffic safety and efficiency and the speed of vehicle as control variable. Starting with the traffic volume condition function, this paper built the traffic speed control model and designed the main lanes speed controller in junction road network.(5) Considering the influence of main lanes and ramps on traffic control, this paper applied the theory and model creation method of intelligent cooperative control to build relevant main lanes multi-ramp integrated control model of junction road network. Finally, it used empirical analysis to prove that considering main lanes and ramps integrated control on the junction road network could resolve or ease the traffic congestion. It could also maintain the stability of the main lanes and improve the overall capacity of the junction road network.更多还原