Data-driven modeling of nonlinear traveling waves

Abstract

Presented is a data-driven machine learning framework for modeling traveling wave spatiotemporal dynamics. The presented framework is based on the steadily propagating traveling wave ansatz, u(x,t)=U(ξ=x−ct+a). For known evolution equations, this coordinate transformation reduces governing partial differential equations to a set of coupled ordinary differential equations (ODEs) in the traveling wave coordinate ξ. Although traveling waves are readily observed in many physical systems, the underlying governing equations may be unknown. For these instances, the traveling wave dynamical system can be modeled empirically with neural ODEs. Presented are these ideas applied to several physical systems that admit traveling waves. Examples include traveling wave fronts, pulses, and wavetrains restricted to one-way wave propagation in a single spatial dimension. Last, applicability to real-world physical systems is presented with an exploration of data-driven modeling of rotating detonation waves.

Document Details

Document Type
Pub Defense Publication
Publication Date
Apr 01, 2021
Source ID
10.1063/5.0043255

Entities

People

  • James Koch

Organizations

  • Air Force Office of Scientific Research
  • University of Texas at Austin

Tags

Readers

  • Artificial Intelligence
  • Atmospheric Science / Meteorology, specifically Wind Wave Turbulence.
  • Calculus or Mathematical Analysis

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference