Answers are grounded in this paper and cite their sources.
Demonstrates an iterative method for solving inverse heat-conduction problems with incomplete boundary info, implying effective temperature recovery.
This paper studies the two-dimensional steady inverse heat-conduction Cauchy problem governed by Laplace’s equation. The temperature field is approximated by a truncated harmonic Trefftz expansion that satisfies the governing equation exactly within the domain, while the boundary conditions are enforced through a collocation procedure. To enhance the stability of the reconstruction, the Kozlov alternating iteration is incorporated, in which the unknown Dirichlet and Neumann data on the inaccessible boundary are updated by solving two well-posed mixed boundary-value problems at each iteration, and the corresponding series coefficients are obtained through a least-squares procedure. Numerical experiments on both smooth and piecewise-smooth geometries demonstrate that the method achieves accurate recovery of temperature and heat flux even in the presence of substantial synthetic noise, with the optimal results typically achieved within a few iterations. The proposed approach is straightforward to implement, does not require meshing, and is well suited for inverse heat-conduction scenarios involving incomplete boundary information.
No takes yet. Share an insight, caveat, or question.
Yan et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: