基于车联网大数据分析的实时路况信息系统关键技术研究

发布时间：2018-04-27 19:15

本文选题：车联网 + 大数据　；参考：《南京邮电大学》2017年硕士论文

【摘要】：随着汽车保有量持续增加,车与路的矛盾日益突出,由此也引发了交通拥堵问题。拥堵问题现在已严重影响了人们的日常出行,而智能交通技术和车联网技术是解决交通拥堵问题的有效途径。以此为背景,本文主要对以车联网大数据为基础的实时路况信息系统相关技术进行研究。通过车联网技术实时采集交通数据并对其进行处理分析,以获得实时准确的路况信息,从而帮助出行人员选择合理的出行路线,实现缓解交通拥堵的目的。本文的主要研究内容如下:(1)研究了基于车联网的实时路况信息获取的“端-管-云”系统架构。首先对该系统的总体架构进行了设计,按照实时交通信息采集、预处理、云平台和数据挖掘技术的顺序进行了层次化功能分析,然后在分析原始交通数据特征的基础上,针对采集的交通信息中存在的数据错误、丢失和冗余等异常现象,制定了相应的清洗规则。(2)针对在实时交通数据采集和清洗的过程中会出现数据缺失的现象,依据交通数据的时空特性,提出了一种动态自适应的数据修复算法。该算法在二次指数平滑法的基础上,将交通数据的空间相关性应用到数据修复中,以提高数据修复质量。仿真实验结果表明,该算法具有更好的数据修复效果。(3)在对道路交通状态识别时,针对交通状态本身的模糊性,本文改进了模糊c均值聚类算法来对海量数据进行聚类分析。改进的算法将模糊c均值聚类算法与Canopy算法结合,用Canopy算法初始化聚类中心与聚类数量,同时引入了Xie-Beni指标来自适应确定模糊加权指数。首先使用改进的算法对交通数据进行聚类并获得道路集群,然后利用交通模式识别算法和交通状态分类标准进行道路交通状态判断。实验表明,该算法有更好的聚类效果,且发生误判的概率更小,能够获得准确的道路交通状态。
[Abstract]:With the increase of vehicle ownership, the contradiction between vehicle and road is becoming more and more serious, which also leads to traffic congestion. The problem of congestion has seriously affected people's daily travel, and intelligent transportation technology and vehicle networking technology is an effective way to solve the problem of traffic congestion. Based on this background, this paper mainly studies the related technology of real-time road condition information system based on big data. The traffic data are collected and analyzed in real time by means of vehicle networking technology to obtain real time and accurate information of road conditions, so as to help travelers choose a reasonable route and achieve the purpose of alleviating traffic congestion. The main contents of this paper are as follows: 1) the "end-tube-cloud" system architecture based on real-time road condition information acquisition based on vehicle networking is studied. Firstly, the overall architecture of the system is designed, and the hierarchical function is analyzed according to the sequence of real-time traffic information collection, preprocessing, cloud platform and data mining technology, and then based on the analysis of the characteristics of the original traffic data. Aiming at the abnormal phenomena such as data error, loss and redundancy in collected traffic information, the corresponding cleaning rules are worked out. According to the temporal and spatial characteristics of traffic data, a dynamic adaptive data repair algorithm is proposed. Based on the quadratic exponential smoothing method, the spatial correlation of traffic data is applied to the data restoration in order to improve the quality of data repair. The simulation results show that the algorithm has better effect of data repair. (3) in view of the fuzziness of the traffic state itself, this paper improves the fuzzy c-means clustering algorithm to analyze the massive data. The improved algorithm combines the fuzzy c-means clustering algorithm with the Canopy algorithm and initializes the clustering center and the number of clusters by using the Canopy algorithm. At the same time, the Xie-Beni index is introduced to determine the fuzzy weighted index adaptively. Firstly, the improved algorithm is used to cluster the traffic data and obtain the road cluster, and then the traffic pattern recognition algorithm and the traffic state classification standard are used to judge the road traffic state. Experiments show that the algorithm has better clustering effect, and the probability of misjudgment is smaller, and the accurate road traffic state can be obtained.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U495;TP311.13

【相似文献】