Research on Intelligent Analysis and Depth Fusion of Multi-Source Traffic Data

Abstract

Intelligence transportation system (ITS) and vehicular networks have attracted the research community in the recent years which generate the “big data” in traffic. However, the collection and application of the big traffic data is limited by the privacy of people who generate data. Besides, data-driven-based ITS only needs information that could reflect one or more types of vehicles at specific intersections, sections, and road networks, rather than that of each individual vehicle. Overall, intelligent analysis and data fusion of multi-source traffic data play an important role to reduce the phenomenon of privacy disclosure and ensure the quality of data. As a result, a complete method of multi-source traffic data analyzing and processing is proposed in this paper, including the data analysis method based on the spatio-temporal regression model and the data fusion method using evidence theory based on the confidence tensor. Finally, the practical data is used to conform the ways proposed before. And not only do the results show that the implicit privacy information has been removed but also present a higher accuracy of the proceed data.

Keywords

• Intelligent transportation system
• multi-source traffic data
• data analysis
• data fusion
• data privacy