Reconstruction and well-posedness of an inverse coefficient problem for a diffusion equation
math.NA, cs.NA, math.AP
Submitted: 2026-11-04
Updated: 2026-11-04
License: http://creativecommons.org/licenses/by/4.0/
The gist: In this work, we investigate the inverse coefficient problem of determining an unknown coefficient in a diffusion equation from a single boundary flux measurement.
Terminology
Abstract
In this work, we investigate the inverse coefficient problem of determining an unknown coefficient in a diffusion equation from a single boundary flux measurement. More precisely, assuming that the domain is cylindrical, we develop a unified theoretical and numerical framework for the stable reconstruction of a space-time-dependent coefficient, by reformulating the inverse problem as a nonlinear equation. The inherent nonlinearity of the problem is transferred to an auxiliary nonlinear equation, enabling the application of techniques and concepts from nonlinear analysis. By combining representation formulas with regularity properties of solutions to diffusion equations, we extend to nonlinear inverse problems an approach that has so far been restricted to linear ones. The theoretical reconstruction procedure is complemented by a numerical analysis of the proposed method, including convergence and error estimates, as well as numerical experiments. The proposed reconstruction algorithm is directly inspired by the theoretical framework.
Sources
Related papers
- Do physics-informed neural networks (PINNs) need to be deep? Shallow PINNs using the Levenberg-Marquardt algorithm
- A Neural-preconditioned Poisson Solver for Mixed Dirichlet and Neumann Boundary Conditions
- Advanced Linear Algebra with Applications - Part I (Numerical linear algebra for PDEs, machine learning, and data assimilation)
- Operator learning meets inverse problems: A probabilistic perspective
- Second-order consistency for learning chaotic dynamics via randomized Jacobian matching
- Windowed thinning and query complexity for the bouncy particle and Zigzag samplers