A Weighted Essentially Non-Oscillatory Numerical Scheme for a Multi-Class LWR Model

Abstract

In this paper we present a high order weighted essentially non-oscillatory (WENO) scheme for solving a multi-class extension of the Lighthill-Whitham-Richards (LWR) model. We first review the multi-class LWR model and present some of its analytical properties. We then present the WENO schemes, which were originally designed for computational fluid dynamics problems and for solving hyperbolic conservation laws in general, and demonstrate how to apply these to the present model. We found through numerical experiments that the WENO method is vastly more efficient than the low order Lax-Friedrichs scheme, yet both methods converge to the same solution of the physical model. It is especially interesting to observe the small staircases in the solution which are completely missed out, because of the numerical viscosity, if a lower order method is used without a sufficiently refined mesh. To demonstrate the applicability of this new, efficient numerical tool, we study the multi-class model under different parameter regimes and traffic stream models. We consider also the convergence of the multi-class LWR model when the number of classes goes to infinity. We show that the solution converges to a smooth profile without staircases when the number of classes increases.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 2002
Accession Number
ADA466239

Entities

People

  • Chi-Wang Shu
  • George C. Wong
  • Mengping Zhang
  • So C. Wong

Organizations

  • University of Science and Technology of China

Tags

DTIC Thesaurus Topics

  • Applied Mathematics
  • Computational Fluid Dynamics
  • Computational Science
  • Computations
  • Convergence
  • Differential Equations
  • Dynamics
  • Eigenvalues
  • Equations
  • Fluid Dynamics
  • Fluid Mechanics
  • Fluids
  • Mathematics
  • Mechanics
  • Partial Differential Equations
  • Physics
  • Traveling Waves

Fields of Study

  • Mathematics

Readers

  • Finite Element Method (FEM) for solving Partial Differential Equations (PDEs)