EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A Physics-Informed Neural Network (PINN) Approach to Over-Equilibrium Dynamics in Conservatively Perturbed Linear Equilibrium Systems

Abhishek Dutta iD, Bitan Mukherjee iD, Sk Aftab Hosen iD, Meltem Turan iD, Denis Constales iD, Gregory Yablonsky iD

Entropy · 20.12.2025 · 28 (1) · 9

DOI: 10.3390/e28010009 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Entropy, online erschienen am 20.12.2025. Verfasst von Abhishek Dutta und 5 weiteren Autor:innen. Beteiligte Einrichtung: Department of Chemical Engineering, Izmir Institute of Technology, Izmir 35430, Turkey. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 23 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Conservatively perturbed equilibrium (CPE) experiments yield transient concentration extrema that surpass steady-state equilibrium values. A physics-informed neural network (PINN) framework is introduced to simulate these over-equilibrium dynamics in linear chemical reaction networks without reliance on extensive time-series data. The PINN incorporates the reaction kinetics, stoichiometric invariants, and equilibrium constraints directly into its loss function, ensuring that the learned solution strictly satisfies physical conservation laws. Applied to three- and four-species reversible mechanisms (both acyclic and cyclic), the PINN surrogate matches conventional ODE integration results, reproducing the characteristic early concentration extrema (maxima or minima) in unperturbed species and the subsequent relaxation to equilibrium. It captures the timing and magnitude of these extrema with high accuracy while inherently preserving total mass. Through the physics-informed approach, the model achieves accurate results with minimal data and a compact network architecture, highlighting its parameter efficiency.

Schlagwörter

Stichwörter der Autor:innen

conservatively perturbed equilibriumcyclic and acyclic mechanismsfinite-time thermodynamicsover-equilibrium dynamicsphysics-informed neural network

Autor:innen und Institutionen

  • Department of Chemical Engineering, Izmir Institute of Technology, Izmir 35430, Turkey.
  • Department of Chemical Engineering, Jadavpur University, Kolkata 700032, India.
  • Department of Mathematics, Ege University, Izmir 35180, Turkey.
  • Department of Electronics and Information Systems, Ghent University, Building S-8, Krijgslaan 281, B-9000 Ghent, Belgium.
  • McKelvey School of Engineering, Department of Energy, Environmental and Chemical Engineering, Washington University in Saint Louis, St. Louis, MO 63130, USA.

ORCID-Profile: Abhishek Dutta, Bitan Mukherjee, Sk Aftab Hosen, Meltem Turan, Denis Constales, Gregory Yablonsky

Zitieren

APA

Dutta, A., Mukherjee, B., Hosen, S. A., Turan, M., Constales, D., & Yablonsky, G. (2025). A Physics-Informed Neural Network (PINN) Approach to Over-Equilibrium Dynamics in Conservatively Perturbed Linear Equilibrium Systems. Entropy, 28(1), 9. https://doi.org/10.3390/e28010009

Vancouver

Dutta A, Mukherjee B, Hosen SA, Turan M, Constales D, Yablonsky G. A Physics-Informed Neural Network (PINN) Approach to Over-Equilibrium Dynamics in Conservatively Perturbed Linear Equilibrium Systems. Entropy. 2025;28(1):9. doi:10.3390/e28010009

BibTeX
@article{Dutta2025,
  author = {Dutta, Abhishek and Mukherjee, Bitan and Hosen, Sk Aftab and Turan, Meltem and Constales, Denis and Yablonsky, Gregory},
  title = {A Physics-Informed Neural Network (PINN) Approach to Over-Equilibrium Dynamics in Conservatively Perturbed Linear Equilibrium Systems},
  journal = {Entropy},
  year = {2025},
  volume = {28},
  number = {1},
  pages = {9},
  doi = {10.3390/e28010009}
}
Kostenlos starten →