EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Progressive Domain Decomposition for Efficient Training of Physics-Informed Neural Network

Dawei Luo, Soo-Ho Jo, Taejin Kim

Mathematics · 04.05.2025 · 13 (9) · 1515

DOI: 10.3390/math13091515 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Mathematics, online erschienen am 04.05.2025. Verfasst von Dawei Luo und 2 weiteren Autor:innen. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Korea government (MOTIE), Scientific Research Foundation of Changzhou College of Information Technology.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study proposes a strategy for decomposing the computational domain to solve differential equations using physics-informed neural networks (PINNs) and progressively saving the trained model in each subdomain. The proposed progressive domain decomposition (PDD) method segments the domain based on the dynamics of residual loss, thereby indicating the complexity of different sections within the entire domain. By analyzing residual loss pointwise and aggregating it over specific intervals, we identify critical regions requiring focused attention. This strategic segmentation allows for the application of tailored neural networks in identified subdomains, each characterized by varying levels of complexity. Additionally, the proposed method trains and saves the model progressively based on performance metrics, thereby conserving computational resources in sections where satisfactory results are achieved during the training process. The effectiveness of PDD is demonstrated through its application to complex PDEs, where it significantly enhances accuracy and conserves computational power by strategically simplifying the computational tasks into manageable segments.

Förderung

  • Korea government (MOTIE)
  • Scientific Research Foundation of Changzhou College of Information Technology

Zitieren

APA

Luo, D., Jo, S., & Kim, T. (2025). Progressive Domain Decomposition for Efficient Training of Physics-Informed Neural Network. Mathematics, 13(9), 1515. https://doi.org/10.3390/math13091515

Vancouver

Luo D, Jo S, Kim T. Progressive Domain Decomposition for Efficient Training of Physics-Informed Neural Network. Mathematics. 2025;13(9):1515. doi:10.3390/math13091515

BibTeX
@article{Luo2025,
  author = {Luo, Dawei and Jo, Soo-Ho and Kim, Taejin},
  title = {Progressive Domain Decomposition for Efficient Training of Physics-Informed Neural Network},
  journal = {Mathematics},
  year = {2025},
  volume = {13},
  number = {9},
  pages = {1515},
  doi = {10.3390/math13091515}
}
Kostenlos starten →