Inverse design of broadband and dispersion-flattened highly GeO₂-doped optical fibers based on neural networks and particle swarm algorithm

Abstract

Reverse design of highly GeO₂-doped silica optical fibers with broadband and flat dispersion profiles is proposed using a neural network (NN) combined with a particle swarm optimization (PSO) algorithm. Firstly, the NN model designed to predict optical fiber dispersion is trained with an appropriate choice of hyperparameters, achieving a root mean square error (RMSE) of 9.47×10⁻⁷ on the test dataset, with a determination coefficient (R²) of 0.999. Secondly, the NN is combined with the PSO algorithm for the inverse design of dispersion-flattened optical fibers. To expand the search space and avoid particles becoming trapped in local optimal solutions, the PSO algorithm incorporates adaptive inertia weight updating and a simulated annealing algorithm. Finally, by using a suitable fitness function, the designed fibers exhibit flat group velocity dispersion (GVD) profiles at 1 400–2 400 nm, where the GVD fluctuations and minimum absolute GVD values are below 18 ps·nm⁻¹·km⁻¹ and 7 ps·nm⁻¹·km⁻¹, respectively.

References

[1] HuangY Z, WangK J, LuoZ Q. Ultrafast Raman fiber laser based on cavity matching scheme and heavily germania-core fiber. Journal of lightwave technology, 2019, 37(13): 2914-2919 J]

[2] YangH L, ZhangR L, JiangX G, et al.. 1.7 µm-1.73 µm tunable ultrafast Raman fiber laser pumped by 1.6 µm dissipative soliton pulses. Optics express, 2022, 30: 45970-45979 J]

[3] ZhuL, WangL L, DongX Y, et al.. Mid-infrared supercontinuum generation with highly germanium-doped silica fiber. Acta optica sinica, 2016, 36(3): 0319001 J]

[4] ChenL, LiaoM S, BiW J, et al.. Coherent supercontinuum generation in step-index heavily Ge-doped silica fibers with all normal dispersion. IEEE photonics journal, 2022, 14(4): 1-6 J]

[5] XuP B, DongY K, ZhouD W, et al.. 1200°C high-temperature distributed optical fiber sensing using Brillouin optical time domain analysis. Applied optics, 2016, 55(21): 5471-5478 J]

[6] WANG P, WANG C C, LI J. Coherent pulse generation based on noise-like-pulse pumped Raman amplification[J]. Optics & laser technology, 2024: 169.

[7] LiP X, WangC C, DuZ Y. The large-mode-area multicore fiber with low and flat normal dispersion profile at 1.5–2.3µm. Optical fiber technology, 2022, 69: 102833 J]

[8] YuQ P, ChangH C. Applications of the finite difference mode solution method to photonic crystal structures. Optical and quantum electronics, 2004, 36(1): 145-163 J]

[9] ShiS Y, ChenC H, PratherD W. Plane-wave expansion method for calculating band structure of photonic crystal slabs with perfectly matched layers. Journal of the optical society of America A, 2004, 21(9): 1769-1775 J]

[10] Da SilvaJ P. Ge-doped microstructured fiber design by genetic algorithm for directional coupling. Microwave and optical technology letters, 2013, 55(2): 281-285 J]

[11] WANG K, TIAN Q H, ZHANG Q, et al. Optimized design of single-pump fiber optical parametric amplifier with highly nonlinear fiber segments using a differential evolution algorithm[J]. Optical engineering, 2019, 58(8).

[12] HUANG S W, YE J W, XU Y, et al. Ultralow confinement loss anti-resonant fibers designed by swarm intelligence[J]. Optics communications, 2023, 530.

[13] ChebaaneS, Ben KhalifaS, JebaliM, et al.. Machine learning based inverse design of raised cosine few mode fiber for low coupling. Optical and quantum electronics, 2024, 56(1): 56 J]

[14] MaZ H, LiY. Parameter extraction and inverse design of semiconductor lasers based on the deep learning and particle swarm optimization method. Optics express, 2020, 28(15): 21971-21981 J]

[15] HeZ Q, DuJ B, ChenX Y, et al.. Machine learning aided inverse design for few-mode fiber weak-coupling optimization. Optics express, 2020, 28(15): 21668-21681 J]

[16] GUO Y H, CHENG Y D, JIANG Y C, et al. Machine learning aided inverse design for flattop beam fiber[J]. Optics communications, 2022, 524.

[17] QinH B, HuangW, SongB B, et al.. Hybrid method for inverse design of orbital angular momentum transmission fiber based on neural network and optimization algorithms. Journal of lightwave technology, 2022, 40(17): 5974-5985 J]

[18] MiaF X, MaY J, LiP L, et al.. Inverse design of broadband dispersion compensation fiber based on deep learning and differential evolution algorithm. IEEE photonics, 2023, 15(3): 7201307[J]

[19] YanQ M, MaR Q, MaY X, et al.. Adaptive simulated annealing particle swarm optimization algorithm. Journal of Xidian University, 2021, 48(4): 120-127[J]

[20] PanX Q, XueL M, LuY, et al.. Hybrid particle swarm optimization with simulated annealing. Multimedia tools and applications, 2019, 78(21): 29921-29936 J]

[21] FlemingJ W. Dispersion in GeO₂-SiO₂ glasses. Applied optics, 1984, 23: 4486-4493 J]

[22] AgrawalG PNonlinear fiber optics, 20135San Francisco, Academic: 27-56 M]

RIGHTS & PERMISSIONS

Tianjin University of Technology