An approach for real-time urban traffic state estimation by fusing multisource traffic data

Abstract

Data fusion is an important tool for estimating urban traffic state when various traffic data are available. In order to get more accurate and comprehensive traffic state, this paper proposes an improved reliability revaluated Dempster- Shafer fusion algorithm (RRDSF) and a framework of real-time traffic state estimation system for fusing multi-source data, tests on the accuracy by real-world traffic data. The framework of real-time traffic state estimation system proposed in this paper shows the feasibility of developing advanced data fusion system for real-time traffic state estimation. The results report in this paper demonstrate that the proposed model can fuse data from loop detectors and probe vehicles to more accurately obtain traffic state estimation than using either of them alone and encourage us to do further work.