Advanced Search
DSpace@MIT

On-line calibration for dynamic traffic assignment

Research and Teaching Output of the MIT Community

Show simple item record

dc.contributor.advisor Moshe E. Ben-Akiva and Haris N. Koutsopoulos. en_US
dc.contributor.author Antoniou, Constantinos en_US
dc.contributor.other Massachusetts Institute of Technology. Dept. of Civil and Environmental Engineering. en_US
dc.date.accessioned 2005-09-27T17:27:29Z
dc.date.available 2005-09-27T17:27:29Z
dc.date.copyright 2004 en_US
dc.date.issued 2004 en_US
dc.identifier.uri http://hdl.handle.net/1721.1/28639
dc.description Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Civil and Environmental Engineering, 2004. en_US
dc.description Includes bibliographical references (p. 149-153). en_US
dc.description.abstract (cont.) application, the EKF has more desirable properties than the UKF. Furthermore, the Limiting EKF provides accuracy comparable to that of the best algorithm (EKF), but with computational complexity which is order(s) of magnitude lower than the other algorithms. en_US
dc.description.abstract In this thesis, an on-line calibration approach for dynamic traffic assignment (DTA) that jointly estimates demand and supply parameters has been developed. The objective of on-line calibration is to introduce a systematic procedure that will use the available data to steer the model parameters to values closer to the realized ones. The approach is general and flexible and imposes no restrictions on the models, the parameters or the data that it can handle. The on-line calibration approach is formulated as a state-space model, comprising transition and measurement equations. A priori values provide direct measurements of the unknown parameters (such as origin-destination flows, segment capacities and traffic dynamics models' parameters), while surveillance information (for example, link counts, speeds and densities) is incorporated through indirect measurement equations. The state vector is defined in terms of deviations of the parameters and inputs that need to be calibrated from available estimates. Standard Kalman Filter theory does not apply to this formulation, as it is not linear. Therefore, three modified Kalman Filter methodologies are presented: Extended Kalman Filter (EKF), Limiting EKF, and Unscented Kalman Filter (UKF). A case study on a freeway network in Southampton, U.K., is used to demonstrate the feasibility of the approach, to verify the importance of on-line calibration, and to test the candidate algorithms. The empirical results from this application support the hypothesis that simultaneous on-line calibration of demand and supply parameters can improve the traffic estimation and prediction accuracy and show significant benefits (over the base case in which only the origin-destination flows are estimated on-line). In this en_US
dc.description.statementofresponsibility by Constantinos Antoniou. en_US
dc.format.extent 153 p. en_US
dc.format.extent 6855363 bytes
dc.format.extent 6876045 bytes
dc.format.mimetype application/pdf
dc.format.mimetype application/pdf
dc.language.iso en_US
dc.publisher Massachusetts Institute of Technology en_US
dc.rights M.I.T. theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. See provided URL for inquiries about permission. en_US
dc.rights.uri http://dspace.mit.edu/handle/1721.1/7582
dc.subject Civil and Environmental Engineering. en_US
dc.title On-line calibration for dynamic traffic assignment en_US
dc.type Thesis en_US
dc.description.degree Ph.D. en_US
dc.contributor.department Massachusetts Institute of Technology. Dept. of Civil and Environmental Engineering. en_US
dc.identifier.oclc 58918362 en_US


Files in this item

Name Size Format Description
58918362.pdf 12.79Mb PDF Preview, non-printable (open to all)
58918362-MIT.pdf 12.78Mb PDF Full printable version (MIT only)

This item appears in the following Collection(s)

Show simple item record

MIT-Mirage