本文選題：交通信息融合　切入點(diǎn)：固定采集　出處：《吉林大學(xué)》2015年碩士論文　論文類型：學(xué)位論文

【摘要】：隨著我國經(jīng)濟(jì)快速的發(fā)展，人口的增加以及收入的增長，汽車的擁有量激增，使得汽車的出行比例越來越高，城市交通面臨著巨大的挑戰(zhàn)。交通擁堵、交通污染、交通安全問題頻繁發(fā)生，嚴(yán)重影響了人們的正常生活，，也對(duì)我國國民經(jīng)濟(jì)的持續(xù)健康發(fā)展提出了挑戰(zhàn)。我國交通的資料比較少，所以很多城市因?yàn)橘Y料的匿乏造成了交通管理、規(guī)劃和控制等各方面的困難。交通管理需要大量的交通數(shù)據(jù)，但是單一的檢測(cè)器獲得的交通數(shù)據(jù)顯然不能滿足它對(duì)交通數(shù)據(jù)的需求。隨著信息融合技術(shù)的誕生并且飛速發(fā)展，這一問題能夠被有效的解決。為了實(shí)現(xiàn)對(duì)交通狀態(tài)進(jìn)行動(dòng)態(tài)估計(jì)，必須首先對(duì)路網(wǎng)絕大多數(shù)（甚至是全部）路段的交通參數(shù)進(jìn)行估計(jì)。然而，目前在我國（甚至是發(fā)達(dá)國家）的高速公路中并不是所有的路段都安裝有檢測(cè)裝置。按照以上所述問題，本文的主要研究內(nèi)容是高速公路多源異構(gòu)交通數(shù)據(jù)融合與預(yù)測(cè)方法，首先對(duì)多源交通信息質(zhì)量評(píng)價(jià)與控制技術(shù)進(jìn)行研究，提出了高速公路中固定采集與移動(dòng)采集的融合方法、基于歷史數(shù)據(jù)與實(shí)時(shí)數(shù)據(jù)的交通流預(yù)測(cè)方法。 固定采集與移動(dòng)采集的融合方法。本文介紹了最小方差加權(quán)平均法的基本原理以及權(quán)值的確定方法。因?yàn)槟壳霸诟咚俟分胁杉煌ㄐ畔⒅饕砸曨l裝置和浮動(dòng)車為主，所以本文以區(qū)間平均速度為例介紹了基于自適應(yīng)加權(quán)平均法的固定采集與移動(dòng)采集的快速融合方法。最后設(shè)計(jì)了融合方法的功能模塊并對(duì)提出的方法進(jìn)行了實(shí)例驗(yàn)證，通過實(shí)驗(yàn)證明了本文提出方法的有效性。 基于歷史數(shù)據(jù)與實(shí)時(shí)數(shù)據(jù)的交通流預(yù)測(cè)方法。由于技術(shù)和管理的限制，高速公路中檢測(cè)器經(jīng)常毀壞或者一些路段根本沒有安裝檢測(cè)器，導(dǎo)致一些路段沒法檢測(cè)交通數(shù)據(jù)，有時(shí)也因?yàn)樘鞖獾脑驅(qū)е掠袡z測(cè)器路段檢測(cè)交通參數(shù)不準(zhǔn)確。本文先利用聚類分析法分析無檢測(cè)器路段以及相鄰的路段的歷史數(shù)據(jù)，利用歷史數(shù)據(jù)和實(shí)時(shí)數(shù)據(jù)對(duì)交通流進(jìn)行預(yù)測(cè)。然后基于交通流理論利用無檢測(cè)器路段的浮動(dòng)車實(shí)時(shí)數(shù)據(jù)和上游路段的歷史數(shù)據(jù)對(duì)交通流進(jìn)行預(yù)測(cè)。最后對(duì)功能模塊進(jìn)行了設(shè)計(jì)并且進(jìn)行了實(shí)驗(yàn)驗(yàn)證。
[Abstract]:With the rapid development of economy in our country, the increase of population and income growth, a surge in the amount of the car, making the car travel more and more, city traffic is facing a huge challenge. Traffic congestion, traffic pollution, frequent traffic safety problems, seriously affecting people's normal life, but also a challenge of China's national economy sustained and healthy development of China's transportation. The data is relatively small, so the lack of data because many of the city caused traffic management, all aspects of planning and control difficulties. Traffic management requires a large amount of data traffic, but the traffic data obtained from single detector cannot satisfy its traffic data the demand of information fusion technology. With the birth and rapid development, this problem can be solved effectively. In order to realize the dynamic estimation of the traffic state, must first vast road network Most (or all) sections of the traffic parameters estimation. However, at present in our country (even developed countries) of the highway and not all sections are installed in a detection device. According to the above mentioned problems, the main content of this paper is the highway of multi-source and heterogeneous traffic data fusion and prediction method research the first technique and control of multi-source traffic information quality evaluation, proposed fusion method of fixed acquisition and mobile acquisition of expressway, prediction method of historical data and real-time data based on traffic flow.
Fusion method of acquisition and fixed mobile acquisition. This paper introduces the method of determining the minimum variance weighted average method and the principle of weight. Because the current highway traffic information collected in the video device and the floating car, so this paper takes segmentspeed as an example to introduce the rapid integration of fixed and mobile data acquisition method of adaptive weighted based on the averaging method. The function module design of the fusion method and the proposed method is verified by experiment, proved the effectiveness of the proposed method.
The prediction method of historical data and real-time data of traffic flow based on technology and management. Due to restrictions in highway detector often destroyed or some sections did not install detector, cause some sections can not detect traffic data, sometimes because of the weather caused a traffic parameter detection detector section is not accurate. This paper first use clustering analysis method the analysis of historical data and the non detector section adjacent to the road, to predict traffic flow by using historical data and real-time data. Then the traffic flow history data by using the theory of non detector section of the floating car data and upstream section on traffic flow prediction based on the end of the function module has been designed and tested.

【學(xué)位授予單位】：吉林大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2015
【分類號(hào)】：U491

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 楊兆升;孫曉梅;王志建;;快速路截面數(shù)據(jù)和車牌識(shí)別數(shù)據(jù)融合算法[J];北京工業(yè)大學(xué)學(xué)報(bào);2009年10期

2 覃頻頻;許登元;黃大明;;基于D-S證據(jù)理論的高速公路事件檢測(cè)信息融合[J];傳感器與微系統(tǒng);2007年04期

3 夏冰,董菁,張佐;周相似特性下的交通流預(yù)測(cè)模型研究[J];公路交通科技;2003年02期

4 劉智勇,李水友;基于信息融合技術(shù)的交通量檢測(cè)方法[J];公路交通科技;2003年02期

5 鄒亮;徐建閩;朱玲湘;溫惠英;;基于浮動(dòng)車移動(dòng)檢測(cè)與感應(yīng)線圈融合技術(shù)的行程時(shí)間估計(jì)模型[J];公路交通科技;2007年06期

6 楊兆升;高學(xué)英;;基于影響因素分類的路段行程時(shí)間融合研究[J];公路交通科技;2010年04期

7 叢玉良;陳萬忠;孫永強(qiáng);陳建;;基于聯(lián)合卡爾曼濾波器的交通信息融合算法研究[J];公路交通科技;2010年07期

8 王立曉;劉鍇;森川高行;;淺析日本融合交通管制和浮動(dòng)車信息的旅行時(shí)間預(yù)測(cè)[J];公路交通科技(應(yīng)用技術(shù)版);2012年08期

9 王立曉;劉鍇;孫小慧;森川高行;;基于多源數(shù)據(jù)融合的動(dòng)態(tài)導(dǎo)航PRONAVI系統(tǒng)[J];公路交通科技(應(yīng)用技術(shù)版);2012年06期

10 胡郁蔥,徐建閩,吳一民,鐘慧玲;基于BP神經(jīng)網(wǎng)絡(luò)的車輛定位融合模型[J];華南理工大學(xué)學(xué)報(bào)(自然科學(xué)版);2004年02期

相關(guān)博士學(xué)位論文 前1條

1 李琦;基于多源數(shù)據(jù)的交通狀態(tài)監(jiān)測(cè)與預(yù)測(cè)方法研究[D];吉林大學(xué);2013年

本文編號(hào)：1650608