基于车牌识别数据的城市道路交通状态判别及旅行时间可靠性的研究

发布时间：2018-05-14 05:37

本文选题：智能交通 + 车牌识别数据　；参考：《青岛科技大学》2017年硕士论文

【摘要】：随着城市规模的扩大和经济的飞速发展,道路交通负荷和交通拥挤现象日益严重,交通供需矛盾日益突出。针对出现的问题,智能交通系统应运而生,它可以提高交通运输效率、改善交通拥挤状况、保障车辆安全运行,为交通运输提供更有效的管理手段,为民众提供更好的服务。近几年,中国许多城市开始采用基于车牌识别的交通信息采集技术来辅助进行道路交通管理。该技术能够得到城市道路行驶车辆的实时监测数据,该数据具有持续生成且数据量大、时间空间相关等特性。目前基于车牌识别数据的智能交通相关的研究较少,高成本的车牌识别系统仅能实现交通监控、流量检测等较为简单的功能,使其性价比较低,因此如何充分利用车牌识别数进行智能交通的相关研究是当前的研究热点。本文首先研究了当前交通信息采集技术、车牌识别系统及其采集的数据特点。针对可能出现的错误数据及缺失数据,给出了相应的识别与修复方法,在数据处理后得出旅行时间的估计值。第二,提出一种基于车牌识别数据的城市道路交通状态判别方法,该方法基于模糊理论实现交通状态判别,并利用青岛市某路段的车牌识别数据对该方法进行实际检验,结果证明该方法确实有效。第三,研究了基于车牌识别数据的旅行时间可靠性分析,发现城市道路旅行时间符合对数正态分布,对对数正态分布函数的性质及旅行时间可靠性的相关指标进行了分析。本文选取缓冲指数作为旅行时间可靠性指标,确定了基于车牌识别数据进行旅行时间可靠性分析的一般流程,通过某路段实际数据进行仿真分析证明了方法的有效性。
[Abstract]:With the expansion of city scale and the rapid development of economy, the traffic load and traffic congestion are becoming more and more serious, and the contradiction of traffic supply and demand is becoming more and more prominent. In view of the problems, Intelligent Transportation system (its) emerges as the times require. It can improve the efficiency of transportation, improve traffic congestion, ensure the safe operation of vehicles, provide more effective management means for transportation and provide better services for the public. In recent years, many cities in China began to use license plate recognition based traffic information collection technology to assist in road traffic management. This technique can obtain real-time monitoring data of urban road vehicles, which has the characteristics of continuous generation, large amount of data, time and space correlation, and so on. At present, there are few researches on intelligent transportation based on license plate recognition data. The high cost license plate recognition system can only realize simple functions such as traffic monitoring, flow detection and so on, which makes its performance-price ratio lower. Therefore, how to make full use of license plate recognition number to carry on the related research of intelligent transportation is the current research hot spot. In this paper, the current traffic information acquisition technology, license plate recognition system and its data collection characteristics are studied. According to the error data and missing data, the corresponding recognition and repair methods are given, and the estimated travel time is obtained after data processing. Secondly, a traffic state discrimination method based on license plate recognition data is proposed. The method is based on fuzzy theory to realize traffic state discrimination, and the method is tested by the license plate recognition data of a section of Qingdao. The results show that the method is effective. Thirdly, the travel time reliability analysis based on license plate recognition data is studied. It is found that urban road travel time conforms to the lognormal distribution. The properties of the logarithmic normal distribution function and the related indexes of travel time reliability are analyzed. In this paper, the buffer index is selected as the travel time reliability index, and the general flow of travel time reliability analysis based on license plate recognition data is determined.
【学位授予单位】：青岛科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491;TP391.41

【参考文献】