HEURISTIC FINITE-STATE MODEL OF THE HUMAN DRIVER IN A CAR-FOLLOWING SITUATION

An approach to modeling human driver behavior in single-lane car following which is based on a finite-state decision structure is considered. The specific strategy at each point in the decision tree was obtained from observations of typical driver behavior. The synthesis of the decision logic is based on position and velocity thresholds and four states defined by regions in the phase plane. The performance of the resulting assumed intuitively logical model was compared with actual freeway data. The match of the model to the data was optimized by adapting the model parameters using a modified PARTAN algorithm. The results indicate that the heuristic model behavior matches actual car-following performance better during deceleration and constant velocity phases than during acceleration periods.

  • Corporate Authors:

    Institute of Electrical and Electronics Engineers

    3 Park Avenue, 17th Floor
    New York, NY  USA  10016-5997
  • Authors:
    • Burnham, G O
    • Bekey, G A
  • Publication Date: 1976-8

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 00153245
  • Record Type: Publication
  • Source Agency: Engineering Index
  • Files: TRIS
  • Created Date: Oct 29 1977 12:00AM