改进节点重要度计算方法在市域公路网规划中的应用

发布时间：2018-05-21 03:24

本文选题：交通运输工程 + 公路网规划　；参考：《长安大学》2014年硕士论文

【摘要】：随着社会经济发展，客运量、货运量与日俱增，对交通基础设施的配置提出了更高的要求。公路网作为交通运输系统的重要组成部分，在满足运输需求方面发挥着重要作用。但是，我国公路网往往与运输需求不匹配，并由此产生诸多问题，为解决这些问题，使公路网更好地服务于所有道路使用者，须深入研究路网规划理论并将理论付诸实践。本论文先从国内外学者在公路网规划方面取得的成果入手，研究公路网的规划思路与布局原则，然后严格按照交通运输部颁行的《公路网规划编制办法》的步骤，对规划区域内节点的相关指标直接进行聚类。本论文为了能够如实反映各指标对其节点在重要程度上的影响，未对各指标分配权重然后线性求和得出重要度，而是采用亲和力传播（AP）算法将节点指标的主成分进行聚类，然后转化为真实指标下的结果。该改进的聚类方法能够有效避免传统节点重要度聚类法因各指标求和而削弱节点差异性的弊端。基于节点聚类结果采用集弹性系数法、曲线拟合法、灰色预测法和神经网络法于一体的组合算法预测交通量，为公路网起讫点的布设及公路的容量规划奠定基础，最后建立公路网评价体系对规划方案进行评价，并采用天水市公路网络规划作为实例，对规划方案的模型进行仿真检验。评价结果表明本论文提出的公路网络规划方案改善了研究区域当前路网的运行质量，并在规划年份能够与运输需求的增长相匹配。此外，本文提出的规划模型易于理解，计算过程易于操作，有良好的普适性，对公路网络规划具有一定的借鉴意义。
[Abstract]:With the development of social economy, passenger and freight traffic are increasing day by day, which puts forward higher requirements for the allocation of transportation infrastructure. As an important part of transportation system, highway network plays an important role in meeting transportation demand. However, the highway network of our country often does not match with the transportation demand, and many problems arise therefrom. In order to solve these problems and make the highway network serve all the road users better, it is necessary to deeply study the theory of road network planning and put the theory into practice. This paper begins with the achievements of scholars at home and abroad in highway network planning, studies the planning ideas and layout principles of highway network, and then strictly follows the steps of "Highway Network Planning compilation method" issued by the Ministry of Transport. The related indexes of the nodes in the planning area are clustered directly. In this paper, in order to reflect the influence of each index on the importance of the node, we do not assign the weight to each index and obtain the importance by linear summation. Instead, we cluster the principal components of the node index by affinity propagation (APP) algorithm. It is then translated into results under real indicators. The improved clustering method can effectively avoid the disadvantage of the traditional node importance clustering method which weakens the node diversity because of the summation of each index. Based on the results of node clustering, the combined algorithm, which integrates elastic coefficient method, curve fitting method, grey forecasting method and neural network method, is used to forecast traffic volume, which lays a foundation for the layout of starting and ending points of highway network and the capacity planning of highway. Finally, the evaluation system of highway network is set up to evaluate the planning scheme, and the model of the planning scheme is tested by simulation using Tianshui highway network planning as an example. The evaluation results show that the proposed highway network planning scheme in this paper improves the operation quality of the current road network in the study area and can match the increase of transportation demand in the planning year. In addition, the planning model proposed in this paper is easy to understand, the calculation process is easy to operate, and has good universality, which is useful for highway network planning.
【学位授予单位】：长安大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U412.12

【参考文献】