Physics-informed neural network with pretraining optimization for ocean acoustic field prediction
The Journal of the Acoustical Society of America · 01.11.2025 · 158 (5) · 3846-3860
DOI: 10.1121/10.0039871 ↗Kurz gefasst
Fachartikel in der Zeitschrift The Journal of the Acoustical Society of America, online erschienen am 14.11.2025. Verfasst von Juncong Tang und 1 weiteren Autor:innen. Beteiligte Einrichtung: State Key Laboratory of Acoustics and Marine Information, Institute of Acoustics, Chinese Academy of Sciences, Beijing 100190, People's Republic of C…. Die Arbeit wurde bisher 14-mal zitiert (Crossref) und nennt 51 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Strategic Priority Research Program of the Chinese Academy of Sciences, Academician Workstation.
Abstract (Originalsprache)
To address the challenge of capturing high-frequency features in ocean acoustic field prediction using Physics-Informed Neural Networks (PINNs), this study introduces an enhanced pretraining optimization approach, termed PreT-OceanPINN, based on a recently proposed OceanPINN framework, a PINN-based method for predicting ocean acoustic pressure fields. By implementing a two-stage strategy—pretraining in a hypothetical environment followed by fine-tuning with real-world data—PreT-OceanPINN significantly improves both prediction accuracy for high-frequency components and training efficiency. In the pretraining stage, the model is trained using envelope signals derived from a simulated environment, allowing it to internalize the underlying physical principles of sound propagation. During the fine-tuning stage, a limited amount of measured data are used to adapt the model to complex real-world conditions. Compared to the standard OceanPINN approach, PreT-OceanPINN delivers more accurate high-frequency predictions without increasing real data dependency, thus demonstrating clear performance advantages. The effectiveness of the proposed method is validated through numerical simulations and experimental data from the SWellEx-96 field experiment.
Autor:innen und Institutionen
- State Key Laboratory of Acoustics and Marine Information, Institute of Acoustics, Chinese Academy of Sciences, Beijing 100190, People's Republic of China.
- University of Chinese Academy of Sciences, Beijing 100049, People's Republic of China.
ORCID-Profile: Haiqiang Niu
Förderung
- National Natural Science Foundation of China
- Strategic Priority Research Program of the Chinese Academy of Sciences
- Academician Workstation
Zitieren
APA
Tang, J., & Niu, H. (2025). Physics-informed neural network with pretraining optimization for ocean acoustic field prediction. The Journal of the Acoustical Society of America, 158(5), 3846-3860. https://doi.org/10.1121/10.0039871
Vancouver
Tang J, Niu H. Physics-informed neural network with pretraining optimization for ocean acoustic field prediction. The Journal of the Acoustical Society of America. 2025;158(5):3846-3860. doi:10.1121/10.0039871
BibTeX
@article{Tang2025,
author = {Tang, Juncong and Niu, Haiqiang},
title = {Physics-informed neural network with pretraining optimization for ocean acoustic field prediction},
journal = {The Journal of the Acoustical Society of America},
year = {2025},
volume = {158},
number = {5},
pages = {3846--3860},
doi = {10.1121/10.0039871}
}
EUVIMED