Statistical model for travel time prediction corresponding to fluctuation of bottleneck capacity

This study expands the previous model to statistically predict the travel time and its variability using data from traffic detector. The travel time variability is presented by a prediction interval that includes the observed travel time at a certain probability. The model is developed to be applicable to a road section, including merging and a junction. Because the congestion in the merging line may decrease the available capacity at the target section, we assume that there is a relationship between the travel time variability and the congestion at other merging line approaching to that section. Instead of directly using the capacity at the bottleneck, this study employs the occupancy to represent the state of the merging line. To validate the model, we conducted an empirical analysis by employing the data observed in Kobe line of the Han-shin Expressway. At the head of the target section, a bottleneck caused by merging exists. The analysis results showed that the proposed model could better estimate the prediction intervals than that of previous model, regardless of the state of the merging line.


  • English

Media Info

  • Pagination: 10p
  • Monograph Title: Transport Research Arena (TRA) 2014 Proceedings

Subject/Index Terms

Filing Info

  • Accession Number: 01530711
  • Record Type: Publication
  • Source Agency: ARRB
  • Files: VTI, TRIS, ATRI
  • Created Date: Jul 23 2014 12:54PM