Inverse design of dispersion for photonic devices based on LSTM and gradient-free optimization algorithms hybridization

Tianhang Yao; Tianye Huang; Bin Yan; Mingfeng Ge; Jie Yin; Chuyu Peng; Lu Li; Wufeng Sun; Perry Ping Shum

doi:10.1364/JOSAB.491490

Journal of the Optical Society of America B
Vol. 40,
Issue 6,
pp. 1525-1532
(2023)
•https://doi.org/10.1364/JOSAB.491490

Inverse design of dispersion for photonic devices based on LSTM and gradient-free optimization algorithms hybridization

Tianhang Yao, Tianye Huang, Bin Yan, Mingfeng Ge, Jie Yin, Chuyu Peng, Lu Li, Wufeng Sun, and Perry Ping Shum

Not Accessible

Your library or personal account may give you access

Get PDF
Email
Share
Get Citation
Copy Citation Text
Tianhang Yao, Tianye Huang, Bin Yan, Mingfeng Ge, Jie Yin, Chuyu Peng, Lu Li, Wufeng Sun, and Perry Ping Shum, "Inverse design of dispersion for photonic devices based on LSTM and gradient-free optimization algorithms hybridization," J. Opt. Soc. Am. B 40, 1525-1532 (2023)

Export Citation
- BibTex
- Endnote (RIS)
- HTML
- Plain Text
Citation alert
Save article

Check for updates

Abstract

As an inherent property of optical devices, dispersion plays an important role in the areas of optical communication and nonlinear optics. Traditional dispersion optimization approaches are time-consuming and power-hungry. In this paper, to accelerate the design of dispersive optical devices, an indirect inverse design method based on the long short-term memory forward model combined with gradient-free optimization algorithms is proposed. In the case of photonic crystal fiber, the results show that the forward model can predict the group velocity dispersion (GVD) with an accuracy of up to 99.62%, and the calculation speed is more than one thousand times faster than the conventional numerical simulations. The prediction accuracy of the inverse model is higher than 93%, with a calculation time of less than 20 s. In the case of slot waveguide, the results show that the forward model can predict the GVD with a prediction accuracy of 96.99% and the inverse design accuracy goes to 99%. The proposed machine learning model offers an efficient tool for dispersion optimization in both fiber and waveguide platforms.

Full Article | PDF Article

More Like This

Multimode waveguide analyses and design based on the FC-LSTM hybrid network

Tianhang Yao, Tianye Huang, Xuming Zeng, Zhichao Wu, Jing Zhang, Dapeng Luo, Xiangli Zhang, Yong Wang, Zhuo Cheng, Xiang Li, Lei Han, and Perry Ping Shum
J. Opt. Soc. Am. B 39(10) 2564-2572 (2022)

Enhanced grey wolf algorithm for automatic tuning of an ensemble neural network in predicting PCF optical properties

Qibo Xu, Hua Yang, Xiaofang Yuan, Jifang Rong, and Jiayi Zhao
Opt. Express 31(26) 43790-43803 (2023)

Hybrid inverse design scheme for nanophotonic devices based on encoder-aided unsupervised and supervised learning

Shuai Yu, Tian Zhang, Jian Dai, and Kun Xu
Opt. Express 31(24) 39852-39866 (2023)

Previous Article Next Article

Data availability

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.

Cited By

You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Figures (9)

You do not have subscription access to this journal. Figure files are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Tables (5)

You do not have subscription access to this journal. Article tables are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

Equations (2)

You do not have subscription access to this journal. Equations are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access Optica Member Subscription

	Min	Max	Step
${Conc}_{0}$	0.05 mol.%	0.50 mol.%	0.05 mol.%
${Conc}_{1}$	${Conc}_{0}$	0.50 mol.%	0.05 mol.%
$r_{1} (0.5 \times d_{1})$	1 µm	5 µm	1 µm
$r_{2} (0.5 \times d_{2})$	1 µm	5 µm	1 µm
$d_{3}$	1 µm	5 µm	1 µm
$λ$	1550 nm	1650 nm	5 nm

	Min	Max	Step
Height	220 nm	340 nm	5 nm
Width	500 nm	800 nm	30 nm
Gap	20 nm	200 nm	–
$λ$	1520 nm	1610 nm	10 nm

	Ge	${L i N b O}_{3}$	Si	${S i}_{3} N_{4}$	${T i O}_{2}$
Air	0	2	4	6	8
${S i O}_{2}$	1	3	5	7	9

	$m_{c}$	Height	Width	Gap
Target	7	270.0 nm	560.0 nm	60.0 nm
Simu-design-I	5	220.0 nm	729.4 nm	33.7 nm
Simu-design-II	6	300.4 nm	604.7 nm	196.1 nm
Simu-design-III	7	305.1 nm	668.8 nm	33.4 nm

	$m_{c}$	Height	Width	Gap
Target	9	310.0 nm	710.0 nm	30.0 nm
Simu-design-I	2	334.6 nm	717.9 nm	27.9 nm
Simu-design-II	3	283.2 nm	756.2 nm	22.3 nm
Simu-design-III	9	220.0 nm	508.8 nm	29.2 nm

Abstract

Data availability

Cited By

Figures (9)

Tables (5)

Equations (2)

Journal of the Optical Society of America B