本文選題：交通工程 + 區(qū)域路網(wǎng)��； 參考：《長(zhǎng)安大學(xué)》2017年碩士論文

【摘要】：面對(duì)機(jī)動(dòng)車保有量增長(zhǎng)速度遠(yuǎn)高于交通基礎(chǔ)設(shè)施建設(shè)速度所帶來(lái)的交通問(wèn)題,利用ITS技術(shù)解決交通癥結(jié)是必然的發(fā)展趨勢(shì),而對(duì)于實(shí)時(shí)和未來(lái)交通流狀態(tài)的準(zhǔn)確把握是發(fā)揮ITS技術(shù)的基礎(chǔ)。城市道路交通系統(tǒng)是一個(gè)復(fù)雜系統(tǒng),具有很強(qiáng)的隨機(jī)性和動(dòng)態(tài)性,同時(shí)具有較強(qiáng)的規(guī)律性和聯(lián)系性。據(jù)此,本文以西安市部分區(qū)域?yàn)槔?以出租車GPS數(shù)據(jù)為基礎(chǔ),研究交通流狀態(tài)識(shí)別方法,挖掘交通流狀態(tài)演化特性,從而掌握實(shí)時(shí)、準(zhǔn)確的交通流運(yùn)行狀態(tài),判別常發(fā)性和偶發(fā)性交通擁堵,為區(qū)域交通管理與控制提供依據(jù),確保城市區(qū)域交通安全、順暢運(yùn)行。基于此,論文對(duì)以下幾個(gè)方面進(jìn)行了研究:(1)交通參數(shù)估計(jì)方法研究本文以大量的出租車GPS數(shù)據(jù)為研究基礎(chǔ)數(shù)據(jù),由于無(wú)法通過(guò)原始數(shù)據(jù)提供的地點(diǎn)速度、GPS時(shí)間、經(jīng)緯度坐標(biāo)、方向、車輛狀態(tài)和數(shù)據(jù)有效性等信息直接準(zhǔn)確的判別交通流狀態(tài),因此提出了利用這些信息估計(jì)交通參數(shù),從而間接獲得更多的有效信息。在對(duì)采集到的出租車GPS數(shù)據(jù)進(jìn)行了預(yù)處理、電子地圖匹配等準(zhǔn)備工作的前提下,構(gòu)建了路段平均車速、路段平均延誤和交叉口平均延誤等交通參數(shù)估計(jì)模型,為后續(xù)的研究工作奠定了基礎(chǔ)。(2)基于SVM的交通流狀態(tài)識(shí)別方法研究對(duì)傳統(tǒng)的SVM二分類算法進(jìn)行改進(jìn),構(gòu)建了基于SVM二叉樹多分類算法的交通流狀態(tài)識(shí)別模型,并以估計(jì)出的三個(gè)交通參數(shù)作為輸入數(shù)據(jù),通過(guò)多次訓(xùn)練標(biāo)定模型參數(shù),實(shí)驗(yàn)證明該模型能夠解決非線性、高維數(shù)的問(wèn)題,實(shí)現(xiàn)對(duì)城市區(qū)域路網(wǎng)的交通流狀態(tài)的準(zhǔn)確識(shí)別。(3)交通流狀態(tài)演化特性分析方法研究以上述研究識(shí)別的交通流狀態(tài)為基礎(chǔ),通過(guò)構(gòu)建基于馬爾可夫模型的交通流狀態(tài)遷移網(wǎng)絡(luò)模型和交通流狀態(tài)演化特性分析模型,挖掘城市區(qū)域路網(wǎng)在一定時(shí)間段內(nèi)交通流狀態(tài)的遷移特性、穩(wěn)定性、偏好性、活躍性、活躍時(shí)間、跳躍遷移、堵塞路段重疊率和時(shí)空分布等特性,從而更加深入地了解城市區(qū)域的整體運(yùn)行規(guī)律,為城市區(qū)域交通管理與控制方案的制定提供依據(jù)。(4)交通流狀態(tài)識(shí)別、演化特性的應(yīng)用研究結(jié)合交通參數(shù)估計(jì)方法、交通流狀態(tài)識(shí)別方法和交通流狀態(tài)演化特性分析方法,以西安市高新區(qū)部分區(qū)域路網(wǎng)、出租車GPS數(shù)據(jù)和路段視頻數(shù)據(jù)為實(shí)驗(yàn)基礎(chǔ),驗(yàn)證本研究構(gòu)建的模型的準(zhǔn)確性、可靠性和實(shí)用性,并分析該區(qū)域內(nèi)路網(wǎng)交通流狀態(tài)的演化過(guò)程及特性,判斷常發(fā)性和偶發(fā)性交通擁堵。
[Abstract]:In the face of the traffic problems caused by the speed of vehicle ownership increasing much faster than the speed of traffic infrastructure construction, it is an inevitable development trend to use its technology to solve the traffic problem. The accurate understanding of the real-time and future traffic flow is the basis of its technology. Urban road traffic system is a complex system, with strong randomness and dynamic, at the same time has strong regularity and connection. On this basis, taking part of Xi'an as an example, based on GPS data of taxis, this paper studies the identification method of traffic flow state, excavates the evolution characteristics of traffic flow state, and grasps the real-time and accurate traffic flow running state. It can provide the basis for regional traffic management and control and ensure the safety and smooth operation of urban traffic. Based on this, this paper studies the following aspects: (1) Traffic parameter estimation method based on a large number of taxi GPS data as the basic data, because the location speed can not be provided through the original data GPS time, latitude and longitude coordinates, The information of direction, vehicle state and data availability can directly and accurately distinguish the traffic flow state. Therefore, it is proposed to use these information to estimate traffic parameters so as to obtain more effective information indirectly. Based on the preprocessing and electronic map matching of the collected taxi GPS data, a traffic parameter estimation model, such as the average speed of the road, the average delay and the average delay at the intersection, is constructed. It lays a foundation for further research work. (2) the traffic flow state recognition method based on SVM improves the traditional SVM two-classification algorithm and constructs a traffic flow state recognition model based on SVM binary tree multi-classification algorithm. The estimated three traffic parameters are used as input data, and the calibration model parameters are trained several times. The experiment shows that the model can solve the nonlinear and high-dimensional problems. To realize the accurate identification of the traffic flow state of the urban road network. (3) the analysis method of the evolution characteristics of the traffic flow state is based on the traffic flow state identified above. By constructing the network model of traffic flow state migration based on Markov model and the analysis model of traffic flow state evolution characteristic, the paper excavates the migration characteristics, stability, preference, activity of traffic flow state of urban regional road network in a certain period of time. The characteristics of active time, jump migration, overlap rate and space-time distribution of blocked road sections, so as to better understand the overall operation law of urban areas, and provide the basis for the formulation of traffic management and control schemes in urban areas. (4) recognition of traffic flow status, The application research of evolution characteristics combined with traffic parameter estimation method, traffic flow state identification method and traffic flow state evolution characteristic analysis method, based on Xi'an High-tech Zone regional road network, taxi GPS data and video data as experimental basis. The accuracy, reliability and practicability of the model are verified, and the evolution process and characteristics of the traffic flow state in the road network are analyzed to judge the frequent and accidental traffic congestion.
【學(xué)位授予單位】：長(zhǎng)安大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2017
【分類號(hào)】：U491

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 何兆成;周亞強(qiáng);余志;;基于數(shù)據(jù)可視化的區(qū)域交通狀態(tài)特征評(píng)價(jià)方法[J];交通運(yùn)輸工程學(xué)報(bào);2016年01期

2 于榮;王國(guó)祥;鄭繼媛;王海燕;;基于支持向量機(jī)的城市道路交通狀態(tài)模式識(shí)別研究[J];交通運(yùn)輸系統(tǒng)工程與信息;2013年01期

3 羅向龍;;基于交通噪聲分析的交通流狀態(tài)識(shí)別方法[J];同濟(jì)大學(xué)學(xué)報(bào)(自然科學(xué)版);2012年12期

4 朱琳;于雷;宋國(guó)華;郭繼孚;溫慧敏;;基于浮動(dòng)車數(shù)據(jù)的交通擁堵時(shí)間維度特征[J];北京交通大學(xué)學(xué)報(bào);2011年06期

5 孫健;劉瓊;彭仲仁;;城市交通擁擠成因及時(shí)空演化規(guī)律分析——以深圳市為例[J];交通運(yùn)輸系統(tǒng)工程與信息;2011年05期

6 劉小明;王灝;李穎宏;王力;張永忠;;基于可拓決策理論的城市交通狀態(tài)識(shí)別研究[J];武漢理工大學(xué)學(xué)報(bào)(交通科學(xué)與工程版);2011年02期

7 祁宏生;王殿海;徐程;;基于合理路徑的擁擠路網(wǎng)交通狀態(tài)分析方法[J];西南交通大學(xué)學(xué)報(bào);2011年01期

8 趙志強(qiáng);張毅;胡堅(jiān)明;;基于GTM-中文版權(quán)目錄TT算法的城市區(qū)域交通狀態(tài)分析[J];吉林大學(xué)學(xué)報(bào)(工學(xué)版);2009年S2期

9 許倫輝;唐德華;;短時(shí)交通流交通狀態(tài)轉(zhuǎn)變及其特性分析[J];系統(tǒng)工程;2009年08期

10 趙志強(qiáng);張毅;李力;胡堅(jiān)明;;基于GTM算法的多維區(qū)域交通狀態(tài)分析[J];中國(guó)公路學(xué)報(bào);2009年04期

相關(guān)碩士學(xué)位論文 前1條

1 倪升華;基于數(shù)據(jù)的交通擁堵評(píng)價(jià)與預(yù)測(cè)方法[D];浙江工業(yè)大學(xué);2014年

，

本文編號(hào)：2056151