01023087

Component

The objective of this paper is to discuss the concept of data fusion for traffic engineering, and to describe a methodological framework that is suitable for combining data from separate sources as well as solutions to this problem. In the second part, the paper presents a case study of estimating travel time by combining synergistically knowledge and information from heterogeneous sources. The first source is conventional traffic detectors data from which one can make an estimation of mean travel time. The second source is probe vehicles reports. Approaching this problem as a typical data fusion problem, we then propose a method for processing it using Dempster-Shafer theory evidence theory. The implementation of this methodology has demonstrated that better results are achieved with fusion than with methods based on individual sources. Besides, it can be shown that the improvement due to the fusion, as measured by correctly classified rates, increases as the degree of precision required of the estimate is increased.

01020180

06-1510

English

18p

2006

Transportation Research Board 85th Annual Meeting

CD-ROM

Figures (3) ; References (58)

Bayes' theorem; Case studies; Data fusion; Detection and identification; Evidence; Information systems; Methodology; Traffic data; Traffic engineering; Travel time

Highways; Law; Operations and Traffic Management; Planning and Forecasting; I72: Traffic and Transport Planning

Transportation Research Board Annual Meeting 2006 Paper #06-1510

TRIS, TRB

Mar 3 2006 10:42AM