本文選題：交通運行狀態(tài)　切入點：車牌跟蹤數(shù)據(jù)　出處：《華南理工大學》2014年碩士論文

【摘要】：城市道路交通運行狀態(tài)的準確實時評價及預測是準確把握城市道路交通系統(tǒng)行為、科學制定交通管理決策的基礎(chǔ)，如何定量描述交通運行狀態(tài)以及如何面向交通參與者對交通狀態(tài)進行不同層次的評價和預測,一直是城市交通研究領(lǐng)域的重點和難點,對這些問題的研究具有重要的理論和現(xiàn)實意義。 本論文以城市交通管理的實際需求為背景,在假定車牌數(shù)據(jù)可以獲取以及對城市道路交通流特性進行分析的基礎(chǔ)上，構(gòu)建起一套適用于城市道路交通系統(tǒng)的完整可操作的交通運行狀態(tài)評價及預測體系，主要研究成果具體體現(xiàn)為以下幾個方面:給出基于出行者心理考慮的交通運行狀態(tài)的定義，并據(jù)此提出了路段單元、OD路徑和路網(wǎng)三個層面上的交通行程指數(shù)作為城市交通運行狀態(tài)的評價指標，驗證了指標的有效性，并對評價單元、更新步長和狀態(tài)劃分標準進行了研究，進而設(shè)計了基于模糊評判的多參數(shù)融合評價模型；對車牌數(shù)據(jù)的處理分析過程進行了研究，針對城市道路單車行程行程時間波動大的特點，設(shè)計了基于時間窗格劃分和數(shù)據(jù)序列分位點值截取的行程時間異常數(shù)據(jù)過濾方法，并對主要數(shù)據(jù)表格進行了設(shè)計；提出了一種基于城市交通特性和時空二維影響因素的短時交通參數(shù)組合預測模型，并用實測的交通流數(shù)據(jù)進行了驗證，結(jié)果表明該模型具有較好的預測精度和對不同交通運行狀態(tài)的適應性；改進了基于預測偏差的交通事件預警方法，加入目標路段單元與上下游主要流入的路段單元的預測偏差值符號判斷的步驟，旨在進一步提高交通事件檢測的準確性；對車牌數(shù)據(jù)的處理分析流程和實時交通運行狀態(tài)評價的可視化進行了仿真及演示。 通過對基于android的手持式交通調(diào)查儀獲取的實測數(shù)據(jù)和基于VISSIM仿真的數(shù)據(jù)進行的處理分析驗證了論文構(gòu)建的城市道路交通運行狀態(tài)評價及預測體系的有效性，，為后續(xù)的相關(guān)研究提供了思路方法。
[Abstract]:The accurate real-time evaluation and prediction of urban road traffic running state is the basis of accurately grasping the behavior of urban road traffic system and making traffic management decision scientifically.How to quantitatively describe the traffic running state and how to evaluate and predict the traffic state at different levels for traffic participants has always been the focus and difficulty in the field of urban traffic research.The study of these problems has important theoretical and practical significance.This paper takes the actual demand of urban traffic management as the background, on the basis of assuming that the license plate data can be obtained and analyzing the characteristics of urban road traffic flow.A set of integrated and operable traffic operation evaluation and prediction system is constructed for urban road traffic system. The main research results are as follows: the definition of traffic running state based on travelers' psychological considerations is given.On the basis of this, the OD path of road section unit and the traffic travel index at three levels of road network are put forward as the evaluation index of urban traffic running state, and the validity of the index is verified, and the evaluation unit is given.The updated step size and state partition criteria are studied, and then a multi-parameter fusion evaluation model based on fuzzy evaluation is designed, and the processing and analysis process of license plate data is studied.Aiming at the large fluctuation of cycle travel time on urban roads, a method of filtering abnormal travel time data based on time pane partition and data sequence locus interception is designed, and the main data tables are designed.A combined forecasting model of short-term traffic parameters based on the characteristics of urban traffic and two-dimensional influence factors of time and space is proposed and verified by the measured traffic flow data.The results show that the model has good prediction accuracy and adaptability to different traffic running states, and improves the traffic event warning method based on forecasting deviation.In order to improve the accuracy of traffic incident detection, the step of judging the prediction deviation value symbol between the target road section unit and the upstream and downstream main section unit is added to further improve the accuracy of traffic incident detection.The simulation and demonstration of the processing and analyzing flow of license plate data and the visualization of real-time traffic running state evaluation are carried out.The validity of the evaluation and prediction system of urban road traffic running state is verified by processing and analyzing the measured data obtained by the hand-held traffic survey instrument based on android and the data based on VISSIM simulation.It provides a way of thinking for the following related research.
【學位授予單位】：華南理工大學
【學位級別】：碩士
【學位授予年份】：2014
【分類號】：U491

【參考文獻】

相關(guān)期刊論文 前9條

1 杭明升,楊曉光,彭國雄;基于卡爾曼濾波的高速道路行程時間動態(tài)預測[J];同濟大學學報(自然科學版);2002年09期

2 聶建強;徐大林;;基于占有率的信號相位交通狀態(tài)實時判別模型[J];指揮控制與仿真;2013年01期

3 陳維榮;關(guān)佩;鄒月嫻;;基于SVM的交通事件檢測技術(shù)[J];西南交通大學學報;2011年01期

4 陳新全,侯志祥,吳義虎,劉振聞;無檢測器交叉口交通流量預測的灰色神經(jīng)網(wǎng)絡模型[J];系統(tǒng)仿真學報;2004年12期

5 張濤;陳先;謝美萍;張sソ

本文編號：1729375