A One-Level O-D Estimation Approach Based on Relaxation

The problem of estimating origin-destination travel demands from limited observations of road traffic conditions has been conventionally formulated as a network design problem (NDP). The intrinsic bi-level structure of NDP poses both analytical and computational difficulties. In an earlier attempt, we overcame the bi-level structure by decoupling equilibrium conditions. The decoupled formulation, however, requires that traffic counts to be available on all links and to satisfy equilibrium conditions perfectly. This paper proposes a path-based formulation which relaxes the equilibrium conditions but still takes users' route choice behavior into account. The relaxation strategy intends to discourage the use of non-equilibrium paths by incorporating Beckmann's cost function into the objective. Both historical O-D information and path travel times measured from probe vehicles are considered as inputs other than traffic counts. An algorithm based on column generation is devised to solve the relaxed formulation and its convergence is proved. Using a benchmark example, we compare the estimation results obtained from bi-level, decoupled and relaxed formulations, and conduct various sensitivity analysis.

Language

  • English

Media Info

  • Media Type: CD-ROM
  • Features: Figures; References; Tables;
  • Pagination: 27p
  • Monograph Title: TRB 86th Annual Meeting Compendium of Papers CD-ROM

Subject/Index Terms

Filing Info

  • Accession Number: 01047029
  • Record Type: Publication
  • Report/Paper Numbers: 07-1122
  • Files: TRIS, TRB
  • Created Date: Feb 8 2007 5:44PM