https://doi.org/10.1140/epjp/s13360-025-06097-3
Regular Article
Novel intelligent neuro-structure optimized Bayesian distributed backpropagation for magnetohydrodynamics flow analysis of double-layer optical fiber coating
1
Department of Electrical Engineering, Islamia University Bahawalpur, 63100, Bahawalpur, Pakistan
2
Department of Electrical and Computer Engineering, COMSATS University Islamabad, Attock Campus, 43600, Attock, Pakistan
3
Future Technology Research Center, National Yunlin University of Science and Technology, 123 University Road, Section 3, 64002, Douliou, Yunlin, Taiwan, ROC
4
Department of Electronic Engineering, Islamia University Bahawalpur, 63100, Bahawalpur, Pakistan
Received:
10
May
2024
Accepted:
5
February
2025
Published online:
27
February
2025
Double-layer coated optical fibers provide vital protection against signal attenuation and mechanical damage, necessitating coatings that offer comprehensive surface coverage to meet stringent mechanical, chemical, and electrical standards. In the current study, a pressure-type die is utilized to coat double-layer optical fibers along with molten polymer, conforming to the Oldroyd 8-constant fluid model. The presented investigation analyzes the influence of magnetohydrodynamic effects during the coating process by leveraging a novel design of intelligent Bayesian regularization scheme (IBRS) to effectively investigate several important physical aspects. Adams numerical solver is employed to solve the associated differential systems, generating reference datasets for a double-layer optical fiber-coated model under various scenarios by variation of wall magnetic parameter, dilatant constant, pseudoplastic constant, and pressure gradient. These parameters play a vital role in enhancing the thickness of coated optical fibers, thereby implying their potential use as controlling parameters for thickness regulation. An intelligent solution strategy is implemented by using supervised artificial neural networks with IBRS. This approach enables immediate numerical approximation outcomes through simulations conducted on training, testing, and validation samples derived from reference datasets of complex geometry. The reliability of the IBRS networks is confirmed through convergence plots depicting mean squared errors (MSEs), effective outputs indicating adaptive control parameters of the optimization algorithm, and histograms based on errors and regression statistics derived from comprehensive simulation studies across several scenarios.
Copyright comment Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
© The Author(s), under exclusive licence to Società Italiana di Fisica and Springer-Verlag GmbH Germany, part of Springer Nature 2025
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
