申请同济大学工学硕士学位论文


基于车牌数据的出行分析

及其在交通领域的应用研究

 

培养单位:交通运输工程学院

一级学科:交通运输工程

二级学科:交通运输规划与管理

研 究 生:刘翀

指导教师:杨东援 教授

二○一二年三月

 

A dissertation submitted to

Tongji University in conformity with the requirements for

the degree of Master of Engineering

Research of Trip Division Method and Application in  

Transportation Field Based on License Plate Data

 

School: School of Transportation Engineering

Discipline: Transportation Engineering

Major: Transport Planning and Management

Candidate: Liu Chong

Supervisor: Prof. Yang Dong-yuan

March, 2012

 

摘  要

交通出行信息是探讨交通需求,进行交通规划、控制管理的基础行为分析数据。传统的人工调查无法获取动态的、大量的个体出行数据,调查工作繁重,消耗大量资源。在智能交通系统蓬勃发展,大量交通信息采集技术被广泛应用的情况下,可以利用信息数据获取交通出行行为。上海市目前在市区内的高架系统主要立交周围和上匝道区域以及部分地面主干道和世博园区周边布置了大量的车牌识别检测器,如何利用数据获取正确的出行信息,是迫切需要解决的研究课题之一

本文以上海市车牌识别系统采集获取的11天原始数据作为研究样本,讨论了如何应用采集数据获取个体机动车出行信息,并对出行信息进行进一步的研究应用的问题。本文的主要研究内容有:(1)对车牌识别数据的数据概况和数据质量进行分析。研究了车牌识别数据的特点、数据质量分析方式、数据总量以及时间空间的分布规律。(2)利用有序聚类的理论,研究利用车牌识别数据划分出行的方法。讨论了几种划分方法的适用性。研究了时间阈值的取值方法。利用上海市的车牌数据进行实例计算,并分析了出行划分的结果。(3)探讨出行信息在交通领域的宏观和微观方面的应用。在宏观层面上,利用车牌识别数据判断了城市机动车的使用程度,用聚类分析方法将使用程度分类。从微观层面上,理论上研究了将直接得到的检测OD转化为断面出行OD的方法。为研究区域内的交通流动方向,发现个体交通的行为规律,辅助交通规划和交通建设提供了理论依据。

关键词: 车牌识别数据,出行划分,机动车使用程度,OD修正

 

ABSTRACT

Traffic trip information is basic behavior analysis data for traffic demand, traffic planning and traffic control. The traditional manual traffic investigation cannot obtain dynamic, substantial individual trip information, and the survey work is so heavy that consume large amounts of resources. In the circumstances of development of intelligent transportation system, which a lot of traffic information collection technology is widely used, traffic trip information can be obtained from data. Currently, a large number of license plate identification detectors are launched on overhead roads, ramps, trunk roads, and around Expo Park in Shanghai. How to use the data to obtain the correct travel trip information is an urgent research topic.

Using the 11-day data from license plate identification system in Shanghai, this paper discusses the method of how to apply the data to access individual motor vehicle travel trip information and take trip information for further research. There are three main contents in this paper. First of all, this paper analyzes the profiles and quality of license plate identification data, including the characteristics of the data, the method of quality analysis and the time space distribution rules. Secondly, using the ordered clustering theory, this paper studies how to divide license plate identification data to obtain trip information; discusses the advantages and disadvantages of several division methods and logical value scale for time threshold. Shanghai license plate data are calculated for trip division as a sample. Third, this paper discusses the application of trip information in the field of transport from macroscopic and microscopic aspects. At the macro level, the license plate identification data can be used to judge the motor vehicle using degree of city. From the microcosmic, the OD of detections can be transformed to OD of road section. In a word, this paper offers a academic basis for the study of regional traffic flow, individual traffic behavior, traffic planning and traffic construction.

Key Words:  license plate identification data, trip division, motor vehicle using degree, OD correction