【摘要】：實(shí)時有效的交通信息采集能為交通誘導(dǎo)、交通擁堵、事故檢測等提供高質(zhì)量的數(shù)據(jù),是城市交通規(guī)劃和交通控制與管理的重要基礎(chǔ)。目前交通信息主要通過線圈檢測器、微波檢測器、視頻檢測器等固定式檢測設(shè)備進(jìn)行采集,存在安裝維護(hù)成本高、檢測范圍有限,難以提供全路網(wǎng)交通信息等不足,已無法滿足智能交通系統(tǒng)對交通信息的需求。RFID技術(shù)憑借其多目標(biāo)快速識別、抗干擾能力強(qiáng)、易部署、成本較低、壽命長、可攜帶車主和車輛信息等優(yōu)勢,在智能交通領(lǐng)域的應(yīng)用越來越廣,是交通信息采集技術(shù)發(fā)展的新方向。通過安裝在路側(cè)的RFID閱讀器與車載電子標(biāo)簽進(jìn)行通信,能夠?qū)⑺熊囕v連入網(wǎng)絡(luò),實(shí)現(xiàn)車聯(lián)網(wǎng),可獲得車牌號、車主信息等靜態(tài)交通信息,這是其他采集技術(shù)所不具備的優(yōu)勢。因此,從理論和技術(shù)層面對基于RFID的車聯(lián)網(wǎng)進(jìn)行全面、深入地研究,具有重要的理論意義和實(shí)用價值。 本文對如何利用RFID技術(shù)構(gòu)建車聯(lián)網(wǎng),以及利用車聯(lián)網(wǎng)進(jìn)行交通信息采集與交通事故自動識別進(jìn)行研究,具體研究內(nèi)容包括： (1)將RFID技術(shù)應(yīng)用于智能交通系統(tǒng)中,并利用RFID技術(shù)構(gòu)建車聯(lián)網(wǎng)模型,針對車聯(lián)網(wǎng)的應(yīng)用特點(diǎn),選取合適的RFID系統(tǒng)參數(shù),為車聯(lián)網(wǎng)的研究提供充分的理論基礎(chǔ)。 (2)研究了車聯(lián)網(wǎng)應(yīng)用過程中的防碰撞問題,并對基于二進(jìn)制樹和基于ALOHA的這兩類標(biāo)簽防碰撞算法進(jìn)行仿真分析,結(jié)合車聯(lián)網(wǎng)的實(shí)際應(yīng)用特點(diǎn),提出了一種適合于車聯(lián)網(wǎng)的標(biāo)簽防碰撞算法,并利用MATLAB模擬采集過程,驗(yàn)證了本文防碰撞算法的效果,為解決車聯(lián)網(wǎng)應(yīng)用過程中的碰撞問題提供良好的底層算法。 (3)給出了交通量、密度、速度等傳統(tǒng)可測參數(shù)和空間占有率、道路行程時間、道路延誤時間、道路選擇概率等車聯(lián)網(wǎng)可測參數(shù)的檢測模型和方法,并利用VISSIM仿真軟件模擬真實(shí)的路網(wǎng)環(huán)境,對車聯(lián)網(wǎng)的交通信息采集過程進(jìn)行模擬,結(jié)果表明基于車聯(lián)網(wǎng)的交通信息采集效果理想。 (4)研究了基于車聯(lián)網(wǎng)的交通事故識別方法,通過分析事故發(fā)生前后交通流參數(shù)的特性,并根據(jù)車聯(lián)網(wǎng)模型所采集到的交通數(shù)據(jù)特點(diǎn),基于排隊論思想提出了一種基于車聯(lián)網(wǎng)數(shù)據(jù)的交通事故自動識別方法,并仿真模擬路網(wǎng)交通事故的發(fā)生,對本文算法的性能進(jìn)行分析。實(shí)驗(yàn)證明,本文算法對事故的檢測效率高,效果良好,且能夠適應(yīng)不同的道路交通狀況。
[Abstract]:Real-time and effective traffic information collection can provide high quality data for traffic guidance, traffic congestion, accident detection and so on. It is an important foundation of urban traffic planning and traffic control and management. At present, traffic information is mainly collected through fixed detection equipment such as coil detector, microwave detector, video detector, etc., which has the disadvantages of high cost of installation and maintenance, limited range of detection, difficulty in providing traffic information of the whole road network, etc. RFID technology can not meet the demand of intelligent transportation system for traffic information. RFID technology has the advantages of fast multi-target identification, strong anti-jamming ability, easy deployment, low cost, long life, and can carry vehicle information, etc. It is more and more widely used in the field of intelligent transportation, which is a new direction of traffic information collection technology. Through the communication between the RFID reader installed on the road side and the electronic tag on the vehicle, all the vehicles can be connected to the network, and the vehicle can be connected to the network, and the static traffic information such as license plate number and owner information can be obtained, which is an advantage that other acquisition technologies do not have. Therefore, it is of great theoretical significance and practical value to conduct a comprehensive and in-depth study on RFID based vehicle networking from the theoretical and technical aspects. This paper studies how to use RFID technology to construct vehicle network and how to use it to collect traffic information and identify traffic accidents automatically. The specific research contents include: (1) applying RFID technology to intelligent transportation system. According to the application characteristics of vehicle networking, the appropriate RFID system parameters are selected to provide a sufficient theoretical basis for the research of vehicle networking. (2) the anti-collision problem in the application process of vehicle networking is studied. The two anti-collision algorithms based on binary tree and Aloha are simulated and analyzed. According to the practical application characteristics of vehicle networking, a label anti-collision algorithm suitable for vehicle networking is proposed, and the acquisition process is simulated by MATLAB. The effectiveness of the anti-collision algorithm in this paper is verified, which provides a good bottom layer algorithm for solving the collision problem in the application of vehicle networking. (3) the traditional measurable parameters such as traffic volume, density, speed and space occupancy are given. The detection model and method of the measurable parameters of vehicle network, such as road travel time, road delay time, road selection probability, etc. The real road network environment is simulated by using Visual IM simulation software, and the traffic information collection process of vehicle networking is simulated. The results show that the traffic information collection based on the vehicle network is effective. (4) the traffic accident identification method based on the vehicle network is studied, and the characteristics of the traffic flow parameters before and after the accident are analyzed. According to the characteristics of traffic data collected by vehicle networking model, a method of automatic identification of traffic accidents based on vehicle network data is proposed based on queuing theory, and the occurrence of road network traffic accidents is simulated. The performance of this algorithm is analyzed. Experimental results show that the proposed algorithm is efficient and effective in accident detection and can adapt to different road traffic conditions.
【學(xué)位授予單位】：北京交通大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2014
【分類號】：U495

【參考文獻(xiàn)】

相關(guān)期刊論文 前2條

1 馬楊;;車路協(xié)同,還有多遠(yuǎn)?[J];中國交通信息化;2011年09期

2 劉建華;楊士航;;淺談車聯(lián)網(wǎng)技術(shù)發(fā)展與應(yīng)用前景[J];中國高新技術(shù)企業(yè);2010年28期

，

本文編號：2120582