多源异构交通数据质量控制关键技术研究

发布时间：2018-01-13 19:15

本文关键词：多源异构交通数据质量控制关键技术研究　出处：《吉林大学》2015年硕士论文　论文类型：学位论文

【摘要】：感知数据、整合集成和分析预测是智能交通总体框架的重要结构。对交通状况及流量的感知采集是感知数据层主要工作，是智能交通整体框架的基石。随着图像处理、信息检测技术不断完善，对于交通状态信息的采集就可以通过磁频、波频、视频和安装在车内GPS等移动定位设备等技术采集，另外基于射频技术和手机信令技术也起到一定补充作用。因而感知层积累的大量基础和运行数据，里面存在着缺失、错误等数据质量问题，若直接使用未被评价过的数据推断当前交通流运行状态，必然会引起与实际交通状态存在差距的误判，可能导致路网更严重的拥堵后果。针对以上现象，本文在分析当前国内外交通信息数据质量有关研究和应用技术的基础上，结合数据质量应用的需求，首先从智能交通数据采集数据特征入手分析，构建基于交通数据的质量评价体系。然后在文中结合多源检测器并行运行的特点，论述了对交通数据质量从多源交通数据准确性、一致性、完整性和实时性逐项评价的方法，提出了交互分析多源信息来评价数据质量的综合方法，并给出清晰的评判流程。在数据质量的控制方面，总结现有的基于空间数据质量控制、基于时间维的数据质量控制技术，提出了基于时空相关的数据质量控制方法，同时也对数据质量的控制流程给出详细的叙述。通过多检测器来源对实际路段的相邻时间、相邻截面、相邻车道采集交通数据，，按上述研究进行了评价和控制方法的验证，由实例分析结果可知，在本文构建的多源异构感知数据质量评价集上基于质量维度的多源交互评价方法和基于时空相关性的数据质量控制方法为交通共用信息平台中多源多维信息感知的实现提供了强有力的技术支撑。
[Abstract]:The perception of data integration and analysis and forecast is an important structure of the overall framework of intelligent transportation. The perception of acquisition of the traffic situation and traffic is the perception data layer mainly work is the cornerstone of the overall framework of intelligent transportation system. With the development of image processing, information detection technology continues to improve, for traffic information collection will be through the magnetic frequency, wave frequency in the car, video and installation of GPS mobile positioning technology in equipment acquisition, RF technology and mobile phone signaling technology also plays a complementary role. Based on a lot of basic operation data and thus the perception layer accumulation, there is a lack of quality problems, errors and other data, if direct use has not been evaluated data to infer the current the traffic flow operation state, will inevitably lead to misjudgment of the gap with the actual traffic status, may lead to more serious consequences. Road network congestion
In view of the above phenomenon, based on the analysis of the current traffic information data quality at home and abroad about the research and application of technology, combined with the data quality needs of the application, first from the analysis of the characteristics of intelligent traffic data acquisition data, constructing quality evaluation system based on traffic data. Then combined with multi detector parallel operation characteristics in this paper, paper the quality of traffic data from multi-source traffic data accuracy, consistency, integrity and real-time itemized evaluation, puts forward a comprehensive method of interactive analysis of multi-source information to evaluate the quality of the data, and given a clear evaluation process.
In the control of the quality of the data, summarize the spatial data quality control based on the existing technology, the quality control data based on the time dimension, and puts forward the method of quality control of spatio temporal correlation based on the data, but also gives the flow control of data quality in detail. Through the adjacent sections of the actual time of multi detector source adjacent section, adjacent lane traffic data collection, according to the above research was conducted to verify the evaluation and control methods, the analysis results from the example, in the perception of multi-source heterogeneous data quality evaluation based on multi set quality dimensions of the interaction based on evaluation method and Realization method of quality control of temporal correlation based on data for the multi-source multi dimension information perception in traffic information platform provides a strong technical support.

【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U495

【参考文献】