A PROTOTYPE SYSTEM FOR REAL-TIME INCIDENT LIKELIHOOD PREDICTION

In the first part of this Innovations Deserving Exploratory Analysis (IDEA) project, efforts were focused on developing and validating the freeway incident likelihood prediction models that form the core of a prototype system for real-time incident likelihood prediction. In the second part of the research, an incident likelihood prediction simulator was developed. This report describes the incorporation of the incident likelihood prediction models into the simulator. A traffic simulator (INTRAS) is used to generate the traffic characteristics (i.e., volume and speed) that are inputs to the freeway incident likelihood prediction models, whereas the environmental conditions are specified by the user. The simulator combines the sequential outputs from an existing incident detection algorithm and those of the incident likelihood prediction models through Bayesian updating.

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00727272
  • Record Type: Publication
  • Contract Numbers: ITS-2
  • Files: TRIS, TRB
  • Created Date: Oct 14 1996 12:00AM