Virtual draw of microstructured optical fiber based on physics-informed neural networks
Optics Express · 11.03.2024 · 32 (6) · 9316
DOI: 10.1364/oe.518238 ↗Kurz gefasst
Fachartikel in der Zeitschrift Optics Express, online erschienen am 29.02.2024. Verfasst von Jinmin Ding und 6 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 50 Literaturangaben. Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
The implementation of microstructured optical fibers (MOFs) with novel micro-structures and perfect performance is challenging due to the complex fabrication processes. Physics-informed neural networks (PINNs) offer what we believe to be a new approach to solving complex partial differential equations within the virtual fabrication model of MOFs. This study, for what appears to be the first time, integrates the complex partial differential equations and boundary conditions describing the fiber drawing process into the loss function of a neural network. To more accurately solve the free boundary of the fiber’s inner and outer diameters, we additionally construct a neural network to describe the free boundary conditions. This model not only captures the evolution of the fiber’s inner and outer diameters but also provides the velocity distribution and pressure distribution within the molten glass, thus laying the foundation for a quantitative analysis of capillary collapse. Furthermore, results indicate that the trends in the effects of temperature, feed speed, and draw speed on the fiber drawing process align with actual fabrication conditions, validating the feasibility of the model. The methodology proposed in this study offers what we believe to be a novel approach to simulating the fiber drawing process and holds promise for advancing the practical applications of MOFs.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Ding, J., Hou, C., Zhao, Y., Liu, H., Hu, Z., Meng, F., & Liang, S. (2024). Virtual draw of microstructured optical fiber based on physics-informed neural networks. Optics Express, 32(6), 9316. https://doi.org/10.1364/oe.518238
Vancouver
Ding J, Hou C, Zhao Y, Liu H, Hu Z, Meng F, et al. Virtual draw of microstructured optical fiber based on physics-informed neural networks. Optics Express. 2024;32(6):9316. doi:10.1364/oe.518238
BibTeX
@article{Ding2024,
author = {Ding, Jinmin and Hou, Chenyang and Zhao, Yiming and Liu, Hongwei and Hu, Zixia and Meng, Fanchao and Liang, Sheng},
title = {Virtual draw of microstructured optical fiber based on physics-informed neural networks},
journal = {Optics Express},
year = {2024},
volume = {32},
number = {6},
pages = {9316},
doi = {10.1364/oe.518238}
}
EUVIMED