Parameter Estimation and Analysis of Car-Following Models

This paper presents a new approach to static and dynamic parameter estimation for car-following models. The approach enables identification of model parameters of general car-following models. Its dynamic nature makes it possible to investigate smoothly or rapidly changing driver behavior as expressed in terms of, for instance changing driver sensitivity or reaction times. From verification using synthetic car following data, the paper shows that the approach is able to track changes in the parameters that describe car-following behavior. The static parameter estimation is relatively straightforward and entails minimization of the difference between predicted and observed behavior. Using vehicle trajectories collected at two motorway sites, (with different traffic conditions), the static approach was applied to test and compare a number of car-following models (including multi-anticipative car-following models, such as the mode of Brexelius). Estimation results showed that the linear Helly Model was the most suitable for the subsequent dynamic estimation, because of its performance as well as its relative simplicity.

Language

  • English

Media Info

  • Media Type: Print
  • Features: Figures; References; Tables;
  • Pagination: pp 245-265
  • Monograph Title: Transportation and Traffic Theory. Flow, Dynamics and Human Interaction. 16th International Symposium on Transportation and Traffic Theory

Subject/Index Terms

Filing Info

  • Accession Number: 01002772
  • Record Type: Publication
  • ISBN: 0080446809
  • Files: TRIS, ATRI
  • Created Date: Aug 3 2005 2:36PM