Feature tracking in physics-informed neural networks via joint optimization of nonlinear deformation manifolds: application to shocks
Abstract
Physics-informed neural networks (PINNs) often converge to inaccurate solutions for conservation laws with shocks, because uniformly distributed collocation points undersample localized features and let the residual be dominated by regions that are already well resolved. We propose a feature-tracking PINN (FT-PINN) in which the solution network is defined on a fixed reference domain and composed with a diffeomorphic deformation map from a parameterized nonlinear manifold. The deformation and solution-network parameters are trained jointly by minimizing the pulled-back conservation-law residual. This lets collocation points concentrate along features of essentially arbitrary geometry, including curved, oblique, and merging shocks, without prior knowledge of their locations. The framework is agnostic to the choice of parameterization. Boundary preservation is enforced exactly through a tangential projection of the displacement, and folding is discouraged by a one-sided penalty on the Jacobian determinant. On four test problems (space-time viscous Burgers with merging shocks, a decelerating Burgers shock, the space-time Euler shock tube, and steady 2D Euler regular shock reflection), FT-PINN resolves shocks at their correct locations with a limited collocation budget. A vanilla PINN with the same architecture, budget, and training either misplaces the shocks or fails to form them.
Related
- Deformation families of open Calabi-Yau manifolds and Steinness
- Membrane Deformation – Finite Elements – Parametric Shape Optimization on the Koch Snowflake
- Event-Triggered Funnel Control for Output Reference Tracking of Unknown Nonlinear Systems
- Saturation on compact manifolds via Stone-Weierstrass type theorem, and forced nonlinear PDEs
- OCP-Based Nonlinear Model Predictive Control for Quadrotor Trajectory Tracking
BibTeX
@misc{thakur2026feature,
title = {{Feature tracking in physics-informed neural networks via joint optimization of nonlinear deformation manifolds: application to shocks}},
author = {Akshay Thakur and Matthew Zahr},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.02230},
eprint = {2610.02230},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.02230},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Feature tracking in physics-informed neural networks via joint optimization of nonlinear deformation manifolds: application to shocks
AU - Akshay Thakur
AU - Matthew Zahr
PY - 2026
DA - 2026/10/01
PB - arXiv
DO - 10.48550/arxiv.2610.02230
UR - https://arxiv.org/abs/2610.02230
AB - Physics-informed neural networks (PINNs) often converge to inaccurate solutions for conservation laws with shocks, because uniformly distributed collocation points undersample localized features and let the residual be dominated by regions that are already well resolved. We propose a feature-tracking PINN (FT-PINN) in which the solution network is defined on a fixed reference domain and composed with a diffeomorphic deformation map from a parameterized nonlinear manifold. The deformation and solution-network parameters are trained jointly by minimizing the pulled-back conservation-law residual. This lets collocation points concentrate along features of essentially arbitrary geometry, including curved, oblique, and merging shocks, without prior knowledge of their locations. The framework is agnostic to the choice of parameterization. Boundary preservation is enforced exactly through a tangential projection of the displacement, and folding is discouraged by a one-sided penalty on the Jacobian determinant. On four test problems (space-time viscous Burgers with merging shocks, a decelerating Burgers shock, the space-time Euler shock tube, and steady 2D Euler regular shock reflection), FT-PINN resolves shocks at their correct locations with a limited collocation budget. A vanilla PINN with the same architecture, budget, and training either misplaces the shocks or fails to form them.
ER -