Evaluation of Real-Time Travel Time Estimation Methods and Disruption Warning Using Noisy Automatic Vehicle Identification Measurements
2009 (English)In: Transportation Research Board Annual Meeting 2009 Paper: Transportation Research Board 88th Annual Meeting, Washington D.C., Jan. 2009, 2009Conference paper (Refereed)
To support the implementation of a real-time traffic information system in Stockholm using automatic vehicle identification (AVI) data, a travel time analysis platform has been developed. The platform is composed of a client-side travel time data analysis program and a database server where historical travel time data are stored. In the study, several existing freeway travel time estimation algorithms and their modifications were implemented in the travel time analysis program, and evaluated using four months data collected on urban streets and arterials in and near the city of Stockholm. The advantages and disadvantages of those algorithms are analyzed using the highly noisy travel time measurements. The results show that with modifications in the parameter settings and the algorithms themselves, the methods are capable of estimating the travel times from the noise-corrupted urban data. In addition to the travel time estimation, the paper presents a simple data disruption recognition approach on the way toward a more comprehensive disruption and incident warning system.
Place, publisher, year, edition, pages
Civil Engineering Computer and Information Science
IdentifiersURN: urn:nbn:se:kth:diva-10902OAI: oai:DiVA.org:kth-10902DiVA: diva2:228431
QC 201102182009-07-312009-07-312011-02-18Bibliographically approved