Physics-informed machine learning for micro-perforated panels: Reproducible prediction, uncertainty, and design optimization
The Journal of the Acoustical Society of America · 01.08.2026 · 160 (2) · 1001-1010
DOI: 10.1121/10.0044535 ↗Kurz gefasst
Fachartikel in der Zeitschrift The Journal of the Acoustical Society of America, online erschienen am 03.08.2026. Verfasst von D J Bainamndi und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Data Science and Artificial Intelligence, Action for Youth and Environment, P.O. Box 379, Maroua, Cameroon. Die Arbeit nennt 33 Literaturangaben.
Abstract (Originalsprache)
Porosity, hole radius, panel thickness, and thickness-to-diameter ratio govern sound absorption in micro-perforated panels (MPPs) through viscous–thermal losses in submillimeter apertures. A physics-informed machine-learning workflow is presented for predicting the frequency-averaged absorption coefficient α¯, combining acoustically motivated feature design with probabilistic prediction and calibrated uncertainty. A quality-controlled dataset of 1000 MPP geometries spanning hole radius, porosity, thickness, perforation-shape class, and frequency-averaged absorption coefficient was used to benchmark 28 regression pipelines under leakage-safe fivefold cross-validation. A smooth Maa-inspired prior, monotone in porosity and saturating in thickness ratio, was combined with residual learning. Gaussian-process (GP) variants were the strongest single learners, with the Matérn-kernel GP and physics-informed (PI) residual GP both achieving root mean square error = 0.138 and R2 = 0.763. Physics-informed Gaussian-process (PI-GP) posteriors provided near-nominal 95% coverage (0.945; mean width = 0.485), whereas split-conformal intervals were more conservative (coverage = 0.970; mean width = 0.573). Measured impedance-tube spectra for six additively manufactured circular MPPs using the Cavity 1 configuration were added as independent validation data. All measured Cavity 1 absorption values fell within the GP and conformal 95% intervals, supporting uncertainty consistency while revealing point-prediction discrepancies associated with unmodeled cavity, fabrication, mounting, and layout effects.
Autor:innen und Institutionen
- Department of Data Science and Artificial Intelligence, Action for Youth and Environment, P.O. Box 379, Maroua, Cameroon.
- Laboratoire Ondes et Milieux Complexes (LOMC), UMR 6294 CNRS, Université Le Havre Normandie, Le Havre, France.
- EDF, DQRE-DIDE-Pôle Cuve, 2 rue Ampère, 93200 Saint-Denis, France.
- Institute for Applied Research in Computing and Engineering (iACE), School of Computing, Engineering and Creative Industries, University of Bedfordshire, Luton LU1 3JU, United Kingdom.
ORCID-Profile: D J Bainamndi, P Maréchal, E Siryabe, R B Tayong
Zitieren
APA
Bainamndi, D. J., Maréchal, P., Siryabe, E., & Tayong, R. B. (2026). Physics-informed machine learning for micro-perforated panels: Reproducible prediction, uncertainty, and design optimization. The Journal of the Acoustical Society of America, 160(2), 1001-1010. https://doi.org/10.1121/10.0044535
Vancouver
Bainamndi DJ, Maréchal P, Siryabe E, Tayong RB. Physics-informed machine learning for micro-perforated panels: Reproducible prediction, uncertainty, and design optimization. The Journal of the Acoustical Society of America. 2026;160(2):1001-1010. doi:10.1121/10.0044535
BibTeX
@article{Bainamndi2026,
author = {Bainamndi, D J and Maréchal, P and Siryabe, E and Tayong, R B},
title = {Physics-informed machine learning for micro-perforated panels: Reproducible prediction, uncertainty, and design optimization},
journal = {The Journal of the Acoustical Society of America},
year = {2026},
volume = {160},
number = {2},
pages = {1001--1010},
doi = {10.1121/10.0044535}
}
EUVIMED