EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Exploring neural network surrogates for high-order mesh-free interpolants

Lucas Gerken Starepravo, Georgios Fourtakas, Steven Lind, Ajay B. Harish, Jack R.C. King

Computational Particle Mechanics · 2026-12 · 18 · 142-156

DOI: 10.1016/j.cpms.2026.09.003 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Computational Particle Mechanics, 2026-12. Verfasst von Lucas Gerken Starepravo und 4 weiteren Autor:innen. Die Arbeit nennt 44 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: The Royal Society.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Mesh-free numerical methods offer flexibility in the discretisation of complex geometries, showing significant potential for problems where mesh-based methods struggle. Although high-order approximations can be obtained through consistency-correction linear systems, such approaches remain prohibitively expensive for Lagrangian formulations, which commonly exhibit low-order convergence. Here we investigate the use of machine learning (ML) to bridge this gap, developing two strategies to couple multilayer perceptrons (MLPs) with the Local Anisotropic Basis Function Method (LABFM) as an exemplar high-order mesh-free method. In the first strategy, neural networks are trained to directly surrogate the high-order kernel; in the second, surrogate models are developed to compute the solution of the dense, low-rank linear systems arising in high-order mesh-free discretisations. The first strategy yields qualitative agreement with validation data but only marginally outperforms inconsistent smoothed particle hydrodynamics (SPH) kernels, with divergent behaviour observed for the Laplacian operator. The second strategy produces solution vectors with mean absolute errors of $\mathcal{O}(10^{-4}$--$10^{-5})$, replicating LABFM second-order convergence rate up to a resolution-dependent limiting accuracy and achieving up to a $5\times$ speedup at equivalent accuracy. However, sensitivity analyses reveal that higher-order approximations impose increasingly stringent accuracy requirements on the predicted solution vector, representing a fundamental challenge for current neural network architectures.

Förderung

  • The Royal Society

Zitieren

APA

Starepravo, L. G., Fourtakas, G., Lind, S., Harish, A. B., & King, J. R. (2026). Exploring neural network surrogates for high-order mesh-free interpolants. Computational Particle Mechanics, 18, 142-156. https://doi.org/10.1016/j.cpms.2026.09.003

Vancouver

Starepravo LG, Fourtakas G, Lind S, Harish AB, King JR. Exploring neural network surrogates for high-order mesh-free interpolants. Computational Particle Mechanics. 2026;18:142-156. doi:10.1016/j.cpms.2026.09.003

BibTeX
@article{Starepravo2026,
  author = {Starepravo, Lucas Gerken and Fourtakas, Georgios and Lind, Steven and Harish, Ajay B. and King, Jack R.C.},
  title = {Exploring neural network surrogates for high-order mesh-free interpolants},
  journal = {Computational Particle Mechanics},
  year = {2026},
  volume = {18},
  pages = {142--156},
  doi = {10.1016/j.cpms.2026.09.003}
}
Kostenlos starten →