城市路网连通可达性矩阵优化模型研究

项目名称： 城市路网连通可达性矩阵优化模型研究

项目编号： No.51208014

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 建筑环境与结构工程学科

项目作者： 周雨阳

作者单位： 北京工业大学

项目金额： 25万元

中文摘要： 近年来主要城市的主干道交通拥堵问题日益突出，严重影响经济发展并成为民生问题的焦点，城市交通路网结构优化迫在眉睫。交通路网是网络的一种特殊形式，运用图论方法研究交通路网的结构，既可以保持交通道路系统和流量系统的物理结构，又便于研究交通系统及其元素之间的相互作用。本研究1.建立交通路网拓扑结构的连通性量模型，及基于交通流量的路网可达性度量模型2.构造路网通达矩阵，建立普适的路网结构均衡优化模型，确定优化目标与约束条件3.通过调查统计、车行试验、公交IC卡、手机信息等多源数据采集融合，提取出行需求及交通流量的信息特征4.根据交通信息特征，分类进行机器学习，运用非负矩阵分解技术求解上述模型，得到交通路网结构优化改造方案。5。根据模型计算结果建立支路微循环交通运营路线，提高道路连通性及公共交通可达性，促进低碳交通出行。6.通过车行试验和仿真实验，验证模型准确性。

中文关键词： 网络结构；矩阵模型；IC卡数据；无人机视频；路径优化

英文摘要： Recently, according to the increasing traffic congestion problem in urban main road network and its serious influence to the economic, the optimization for the city road network is imminently taken into account,since it is also the focus of people's livelihood. The graph theory can be intruduced to study the traffic network structure, as it is a special form of the normal network. The graph theory can not only keep the physical character of the road system, but also be facilitated to illuminate the correlation of the traffic system elements. In this research, 1.two measurement models of connectivity and accessibility are proposed; 2.the connectivity and accessibility matrix model is founded to solve the road network network optimization problem with the objective function and several restrictions; 3. the traffic information character extraction is processed in the multi-source database, such as the investigation, driving test, the information from the GSM and the IC card system etc. 4. according to the transport information characters, the model is solved using non-negative matrix factorization technique based on the machine learning on several sub training sets; 5.the optimation model and its solution results lay the foundation of the micro-cycle operation route planning with better connectivity and accessib

英文关键词： network structure；matrix model；smar card data；drone video；route optimization