节点文献

城市快速路交通流采集与控制相关问题研究

Study on Traffic Flow Collection and Control of Urban Express Ways

【作者】 王亮

【导师】 王正欧;

【作者基本信息】 天津大学 , 管理科学与工程, 2005, 博士

【摘要】 快速路是城市道路的最高层次。城市快速路交通流信息采集、融合、交通控制等问题是快速路智能交通管理系统建设中的关键问题。本论文研究了城市快速路交通流采集与控制的相关问题,主要内容如下:1.针对快速路同一检测面上多检测器的交通数据融合问题,论文提出了两种交通数据融合方法,一种是基于NN-FR的数据融合方法,具体为在设计ANFIS置信度判别器基础上,建立了自适应NN-FR信息融合系统。另一种是基于支持向量机的交通数据融合方法,具体采用支持向量机回归理论实现数据融合。论文还对这两种方法做了比较研究,并对某市的某一检测面上的三种检测器所测量的交通数据的融合进行了仿真计算。2.针对快速路交通中不同检测面数据的关联性问题,论文提出了两种不同检测面数据关联性的确定方法。一方面,针对交通数据的非平稳性,提出了一种基于协整理论的不同检测面交通流量数据关联性的确定方法。另一方面,针对交通数据所具有的数据量庞大、有噪声、随机、有复杂的数据结构等特性,提出了一种基于时间序列相似性分析的不同检测面交通流量数据关联性的确定方法。在对交通中不同检测面数据进行关联性研究基础上,论文对检测器布点优化以及检测器状态校验进行了初步探讨研究。3.针对快速路交通控制问题,提出了一种基于递阶协调控制的快速路控制求解算法。在详尽分析适合于快速路交通控制主要方法的基础上,结合大系统递阶控制思想,将快速路整体控制分解为对单一路段的控制,并通过协调层调整各路段的控制策略,以满足道路系统总体的性能要求。此外,对原交通流微观仿真系统进行了改造,使之适应快速路交通控制仿真的需要,并利用该仿真系统对本文提出的控制方法,进行了仿真。4.讨论并建立了快速路交通管理系统框架结构。论文作者结合工作实际,分析了城市快速路交通管理的特点和功能需求,给出了城市快速路管理系统的基本框架结构,为交通管理系统的建立和实施打下了基础。

【Abstract】 Express way is the highest level of urban roads. Problems including informationcollection of traffic flow, data fusion and traffic control are the key problems inconstructing ITS for urban express ways. This dissertation studies on the problemsabout information collection of traffic flow and traffic control of urban express ways.The main points of this dissertation are as follows:Aimed at problem of data fusion which is collected by multi-detector in the samedetecting section in express ways, this dissertation presents two methods. One is datafusion based on Neural Network and Fuzzy Reasoning (NN-FR). On basis ofdesigning ANFIS confidence estimator, an adaptive NN-FR data fusion system is setup. The other one is the method based on support vector machine (SVM), which isachieved by SVM regression. The two methods are compared and are illustrated bytraffic data obtained from three types of detectors in a detecting section of a certaincity.Aimed at problem of correlation of different detecting sections in express ways,this dissertation proposes two methods. On one hand, according to the unstationaryof traffic data, a method based on cointegration theory is presented. On the otherhand, because of the traffic data being of numerous, noisy, and complicated, amethod based on similarity analysis of time series is brought forward. On the basisof studying on the correlation of different detecting sections, the allocationoptimization and the status checking of detectors are analyzed primarily.An algorithm using hierarchical coordinate control is proposed to solve theexpress ways control problem. After analyzing the main control methods suitable forexpress ways control, control of the whole express way system is divided into thecontrols of several segments, which are coordinated by the coordinate level to satisfythe performance request of the system, according to the theory of hierarchical controlmethod. Then some changes are made to the previous traffic flow microcosmicsimulation system to make it adaptable for the simulation of express way. After that asimulation experiment is implemented with the proposed control method.The framework of traffic management system of express ways is discussed andbuilt. The dissertation analyzes the characters and functions requirement of trafficmanagement system of express ways according to the authors work experience. As aresult, the framework of traffic management system of express ways is given, whichmakes a foundation for building and actualization of the traffic management system.

【关键词】 快速路数据融合关联性交通控制
【Key words】 express waydata fusioncorrelationtraffic control.
  • 【网络出版投稿人】 天津大学
  • 【网络出版年期】2006年 08期
节点文献中: