【作者】 聂伟；

【导师】 邵春福；

【作者基本信息】 北京交通大学 ， 交通运输规划与管理， 2007， 博士

【摘要】 都市圈是城市发展的必然阶段，是国家和区域竞争力的重要体现。而道路系统是都市圈内一切实物流动的重要载体，是引导和促进都市圈经济和社会持续、快速、协调发展的主要动力。为了有效地满足都市圈内产生的交通需求，避免建设资源和道路设施的浪费，必须制定科学、系统的道路网络络规划。随着都市圈社会经济和交通运输的发展，对道路网络系统的要求不断提高，与此同时，以往道路规划和建设中存在的问题不断显现，使得对道路网络优化理论的深入研究尤显迫切。本文以最优化理论和现代组合优化理论为基础，以系统工程的思想为指导，借鉴已有研究成果，运用系统分析、定性与定量研究相结合等多种方法，对道路网络优化及其评价的理论和方法进行了深入分析和研究。本文首先介绍了都市圈的基础理论，利用改进的引力和场强模型，对北京都市圈范围界定进行实证性研究；从北京都市圈的现实条件和功能“扩散”、“再集中”的理念出发，提出了北京都市圈的“环型多功能集约型”整体发展思路。其次，对都市圈道路交通发展趋势进行了分析和预测，利用区位商模型分析了都市圈的道路交通及其客货运输区位的变化趋势；建立等长度递补变维分形模型预测了都市圈的道路客、货运交通结构的发展趋势；利用信息熵理论，建立了都市圈道路交通系统熵的测算模型，对北京都市圈道路交通系统演化方向进行了分析。再次，在交通需求预测的四阶段方法的基础上，根据都市圈内城际交通的特点，利用经济性、快速性和方便性指标，建立适合都市圈客运交通方式选择的Logit模型，确定模型参数的标定方法，并以北京都市圈为例，进行了都市圈道路交通需求预测的实证性研究。然后，在已有的交通网络设计研究成果的基础上，根据道路网络优化的实际需要，建立了道路网络优化的两阶段双层优化模型——远期优化模型和近期优化模型，研究了道路网络设计问题双层模型的遗传算法解法，并以北京都市圈道路网络优化问题为对象进行了实证性研究。最后，提出了基于DEA和加权灰色关联分析的道路网络评价方法，并将该组合方法用于评价北京都市圈道路网络综合发展水平。这种方法综合了DEA和灰色关联分析两种方法的优点，对每一个评价对象都去寻求最有利的指标权重系数分配，使得每一个评价对象对于标准对象的关联度都是相对最优的，从而得到客观公正的评价结果。论文取得的主要研究成果包括：改进了引力和场强模型，探讨了北京都市圈“环形多功能集约型”发展思路；建立了交通结构发展趋势预测的等长度递补变维分形模型和道路交通系统发展趋势预测的熵变模型；以小汽车、长途汽车和城际铁路列车为研究对象，研究了适合都市圈客运交通方式选择的Iogit模型；提出了道路网络优化的两阶段双层模型——远期规划模型和近期规划模型，并研究了模型的遗传算法求解方法；将数据包络分析方法(DEA)和灰色理论相结合，建立了都市圈道路网络综合评价的基于DEA的加权灰色关联分析模型。更多还原

【Abstract】 Metropolitan area is the inevitable stage for city, and it is the important embodiment of competitiveness for a country or region. Road system is an important carrier for all material and objects flowed in metropolitan area, which leads the economy and society of the region to a continuous, fast and harmonious development.In order to meet with the traffic demand in the metropolitan area effectively, and to avoid the waste of construction sources and road facilities, it is required to establish scientific and systematic road network planning. With the development of the social economy and communications and transportation, Traffic demand on road network system improves constantly. Meanwhile, a lot of problems concerning to road planning and construction in the past appears, so it is especially urgent to study road network planning theory deeply. Road network optimization theory and evaluation method were studied in the dissertation on the basis of optimization theory and modern combination optimization theory, thought of systematic engineering, existed findings about road network optimization, systems analysis, qualitative analysis and quantitative analysis, and so on.Firstly, basic theory of the metropolitan area was introduced in the dissertation, and an improved gravitation and field intensity model was used to process demonstration research on definition of Beijing metropolitan area scope. From the actual condition of Beijing and the thought of "Diffusion" and "Re-concentration", a whole development idea of round and multi-functional intensive structure for Beijing metropolitan area was put forward.Secondly, trend of road network system of metropolitan area was analyzed and forecasted. A location quotient model was used to analyze the tendency of the location quotient of each cities and its passenger transportation and freight transportation, an Equal-length and Changeable-dimension Fractal Model was set up to predict the development tendency of road passengers and goods traffic structure, and an entropy change model was established to analyze the evolution direction of road network system of Beijing metropolitan area based on entropy theory.Thirdly, on the basis of the "four-step" method for traffic demand forecasting, and according to the characters of intercity traffic in metropolitan area, three indices, including economy, rapidity and convenience, were used to set up a new logit model suitable to modal split of passenger traffic, and method for calibrating the parameters of the model was introduced. Next, Beijing metropolitan was taken as an example to carry through the demonstrative study of traffic demand forecasting.Fourthly, according to the actual needs of the road network optimization, a two-stage bi-level programming model was presented in the dissertation on the basis of existed research of network design, including long-term optimization model and short-term optimization model. And Genetic Algorithm was used to solve the bi-level models, and the road network optimization of Beijing metropolitan area was taken as an example.Finally, theories of DEA and weighted gray correlation analysis were combined to evaluate the comprehensive level of road network system in metropolitan area. The new method can avoid the subjectivity when the weight vector is determined and get the optimality of the prior order evaluation by non-uniform weight.The following are the main conclusions of this dissertation:(1) A model of gravity and field strength was improved, and a comprehensive development strategy of "round and multi-functional intensive structure" was presented.(2) A new fractal forecasting model was put forward-"Equal-length and Changeable-dimension Fractal Model Using Compensating Figure", and an entropy change model was build to analyze the developing tendency of road system.(3) Using automobile, long-distance bus and railway as objects, a new logit model was presented, which is more suitable to modal spit in metropolitan area.(4) A two-stage and bi-level programming model for mixed transportation network design problem was presented which is composed of long-term model and short-term model. And the Genetic Algorithm is used to solve the model.(5) A new model called weighted gray correlation analysis method based on DEA (Data Envelopment Analysis) was presented to evaluate road network system in metropolitan area.更多还原