Reinforcement Operator Learning (ROL): A hybrid DeepONet-guided reinforcement learning framework for stabilizing the Kuramoto–Sivashinsky equation
PLOS One · 30.01.2026 · 21 (1) · e0341161
DOI: 10.1371/journal.pone.0341161 ↗Kurz gefasst
Fachartikel in der Zeitschrift PLOS One, online erschienen am 30.01.2026. Verfasst von Nadim Ahmed und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Physical Sciences, Independent University, Bangladesh, Dhaka, Bangladesh. Hauptthemen laut MEDLINE: Neural Networks, Computer. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 57 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
This study presents Reinforcement Operator Learning (ROL)—a hybrid control paradigm that marries Deep Operator Networks (DeepONet) for offline acquisition of a generalized control law with a Twin-Delayed Deep Deterministic Policy Gradient (TD3) residual for online adaptation. The framework is assessed on the one-dimensional Kuramoto–Sivashinsky equation, a benchmark for spatio-temporal chaos. Starting from an uncontrolled energy of 42.8, ROL drives the system to a steady-state energy of 0.40 ± 0.14, achieving a 99.1% reduction relative to a linear–quadratic regulator (LQR) and a 64.3% reduction compared with a pure TD3 agent. DeepONet attains a training loss of 7.8 × 10 −6 after only 200 epochs, enabling the RL phase to reach its reward plateau 2.5 × sooner and with 65% lower variance than the baseline. Spatio-temporal analysis confirms that ROL restricts state amplitudes to ± 1.8 —three-fold tighter than pure TD3 and an order of magnitude below LQR—while halving the energy in 0.19 simulation units (33% faster than pure TD3). These results demonstrate that combining operator learning with residual policy optimisation delivers state-of-the-art, sample-efficient stabilisation of chaotic partial differential equations and offers a scalable template for turbulence suppression, combustion control, and other high-dimensional nonlinear systems.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- Department of Physical Sciences, Independent University, Bangladesh, Dhaka, Bangladesh.
- Department of Arts and Sciences, Ahsanullah University of Science and Technology (AUST), Dhaka, Bangladesh.
- Department of Mathematics, American International University-Bangladesh, Dhaka, Bangladesh.
- Department of Applied Mathematics, University of Dhaka, Dhaka, Bangladesh.
- Department of Information and Computer Science, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia.
- SDAIA-KFUPM Joint Research Center for AI, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia.
ORCID-Profile: Muhammad Sajjad Hossain, Md Fayz-Al-Asad, Md Awlad Hossain
Zitieren
APA
Ahmed, N., Ashraful Babu, M., Hossain, M. S., Fayz-Al-Asad, M., Awlad Hossain, M., Mortuza Ahmmed, M., Mostafizur Rahman, M., & Mahmud, M. (2026). Reinforcement Operator Learning (ROL): A hybrid DeepONet-guided reinforcement learning framework for stabilizing the Kuramoto–Sivashinsky equation. PLOS One, 21(1), e0341161. https://doi.org/10.1371/journal.pone.0341161
Vancouver
Ahmed N, Ashraful Babu M, Hossain MS, Fayz-Al-Asad M, Awlad Hossain M, Mortuza Ahmmed M, et al. Reinforcement Operator Learning (ROL): A hybrid DeepONet-guided reinforcement learning framework for stabilizing the Kuramoto–Sivashinsky equation. PLOS One. 2026;21(1):e0341161. doi:10.1371/journal.pone.0341161
BibTeX
@article{Ahmed2026,
author = {Ahmed, Nadim and Ashraful Babu, Md and Hossain, Muhammad Sajjad and Fayz-Al-Asad, Md and Awlad Hossain, Md and Mortuza Ahmmed, Md and Mostafizur Rahman, M and Mahmud, Mufti},
title = {Reinforcement Operator Learning (ROL): A hybrid DeepONet-guided reinforcement learning framework for stabilizing the Kuramoto–Sivashinsky equation},
journal = {PLOS One},
year = {2026},
volume = {21},
number = {1},
pages = {e0341161},
doi = {10.1371/journal.pone.0341161}
}
EUVIMED