A STATISTICAL TEST FOR MEASURING THE PERFORMANCE OF TRAFFIC ASSIGNMENT MODELS

This paper was presented at Session 28-Transport Planning Techniques 2. The paper briefly describes some research which was aimed at demonstrating that available statistical tests could be used to accurately measure the performance of traffic assignment models. The tests that have been used to quantitatively measure the performance of traffic assignment models are first described and critically examined. On the basis of this examination it is shown that the Kolmogorov- Smirnov two-sample test is superior to all others. The performance of the traffic assignment model used in the Sydney area transportation study, and the performance of those of the Western Australian Main Roads Department and Wilbur Smith and Associates Transport Planning Packages, are measured using the Kolmogorov-Smirnov and other tests. On the basis of the results of these tests and the findings of other related research it is concluded that multi-path and multiple route traffic assignment models and the incremental capacity restraint models based on them are the most accurate. They are the ones that are more likely to give accurate results for both macro and micro scale planning. (TRRL)

Media Info

  • Features: References; Tables;
  • Pagination: p. 143-152
  • Serial:
    • Volume: 9
    • Issue Number: 6

Subject/Index Terms

Filing Info

  • Accession Number: 00310134
  • Record Type: Publication
  • Source Agency: ARRB
  • Files: ITRD, TRIS, ATRI
  • Created Date: Jul 22 1980 12:00AM