| ELECTROMAGNETISM, OPTICS, ACOUSTICS, HEAT TRANSFER, CLASSICAL MECHANICS, AND FLUID DYNAMICS |
Prev
Next
|
|
|
Inverse design of 3D integrated high-efficiency grating couplers using deep learning |
| Yu Wang(王玉)1,2, Yue Wang(王越)2, Guohui Yang(杨国辉)3, Kuang Zhang(张狂)3, Xing Yang(杨星)4,5, Chunhui Wang(王春晖)2,†, and Yu Zhang(张雨)4,5,‡ |
1 The College of Computer and Control Engineering, Northeast Forestry University, Harbin 150040, China;
2 National Key Laboratory of Laser Spatial Information, Harbin Institute of Technology, Harbin 150001, China;
3 The School of Electronic and Information Engineering, Harbin Institute of Technology, Harbin 150001, China;
4 The State Key Laboratory of Pulsed Power Laser Technology, College of Electronic Engineering, National University of Defense Technology, Changsha 410073, China;
5 Jianghuai Advance Technology Center, Hefei 230037, China |
|
|
|
|
Abstract In recent years, the use of deep learning to replace traditional numerical methods for electromagnetic propagation has shown tremendous potential in the rapid design of photonic devices. However, most research on deep learning has focused on single-layer grating couplers, and the accuracy of multi-layer grating couplers has not yet reached a high level. This paper proposes and demonstrates a novel deep learning network-assisted strategy for inverse design. The network model is based on a multi-layer perceptron (MLP) and incorporates convolutional neural networks (CNNs) and transformers. Through the stacking of multiple layers, it achieves a high-precision design for both multi-layer and single-layer raster couplers with various functionalities. The deep learning network exhibits exceptionally high predictive accuracy, with an average absolute error across the full wavelength range of 1300-1700 nm being only 0.17%, and an even lower predictive absolute error below 0.09% at the specific wavelength of 1550 nm. By combining the deep learning network with the genetic algorithm, we can efficiently design grating couplers that perform different functions. Simulation results indicate that the designed single-wavelength grating couplers achieve coupling efficiencies exceeding 80% at central wavelengths of 1550 nm and 1310 nm. The performance of designed dual-wavelength and broadband grating couplers also reaches high industry standards. Furthermore, the network structure and inverse design method are highly scalable and can be applied not only to multi-layer grating couplers but also directly to the prediction and design of single-layer grating couplers, providing a new perspective for the innovative development of photonic devices.
|
Received: 13 April 2025
Revised: 26 June 2025
Accepted manuscript online: 01 August 2025
|
|
PACS:
|
41.85.-p
|
(Beam optics)
|
| |
42.25.-p
|
(Wave optics)
|
| |
84.71.Mn
|
(Superconducting wires, fibers, and tapes)
|
|
| Fund: This study was sponsored by the National Key Scientific Instrument and Equipment Development Projects of China (Grant No. 62027823) and the National Natural Science Foundation of China (Grant No. 61775048). |
Corresponding Authors:
Chunhui Wang, Yu Zhang
E-mail: wangch_hit@163.com;zhangyu24d@nudt.edu.cn
|
Cite this article:
Yu Wang(王玉), Yue Wang(王越), Guohui Yang(杨国辉), Kuang Zhang(张狂), Xing Yang(杨星), Chunhui Wang(王春晖), and Yu Zhang(张雨) Inverse design of 3D integrated high-efficiency grating couplers using deep learning 2026 Chin. Phys. B 35 024101
|
[1] Blumenthal D J, Heideman R, Geuzebroek D, Leinse A and Roeloffzen C 2018 Proc. IEEE 106 2209
[2] Rahim A, Ryckeboer E, Subramanian A Z, et al. 2017 J. Light. Technol. 35 639
[3] Shi Y, Zhang Y, Wan Y, Yu Y, Zhang Y, Hu X, Xiao X, Xu H, Zhang L and Pan B 2022 Photonics Res. 10 A106
[4] Siew S Y, Li B, Gao F, et al. 2021 J. Light. Technol. 39 4374
[5] Sun C, Yang L, Li B, Shi W, Wang H, Chen Z, Nie X, Deng S, Ding N and Zhang A 2021 Opt. Lett. 46 5699
[6] Xiang C, Jin W and Bowers J E 2022 Photonics Res. 10 A82
[7] Boes A, Corcoran B, Chang L, Bowers J and Mitchell A 2018 Photonics Rev. 12 1700256
[8] Churaev M, Wang R N, Riedhauser A, et al. 2023 Nat. Commun. 14 3499
[9] Xiang C, Guo J, Jin W, et al. 2021 Nat. Commun. 12 6650
[10] Goyvaerts J, Kumari S, Uvin S, Zhang J, Baets R, Gocalinska A, Pelucchi E, Corbett B and Roelkens G 2020 Opt. Express 28 21275
[11] Jacob B, Camarneiro F, Borme J, Bondarchuk O, Nieder J B and Romeira B 2022 ACS Appl. Electron. Mater. 4 3399
[12] Subramanian A Z, Ryckeboer E, Dhakal A, et al. 2015 Photonics Res. 3 B47
[13] Scheerlinck S, Schrauwen J, Laere F V, Taillaert D, Thourhout D V and Baets R 2007 Opt. Express 15 9625
[14] Zhang J, Yang J, Lu H, Wu W, Huang J and Chang S 2015 Chin. Opt. Lett. 13 091301
[15] Papes M, Cheben P, Benedikovic D, et al. 2016 Opt. Express 24 5026
[16] Maire G, Vivien L, Sattler G, et al. 2008 Opt. Express 16 328
[17] Wang P, Luo G, Xu Y, et al. 2020 Photonics Res. 8 912
[18] Mak J C, Wilmart Q, Olivier S, Menezo S and Poon J K 2018 Opt. Express 26 13656
[19] Yoon J, Kim J Y, Kim J, Yoon H, Neeli B, Park H H and Kurt H 2023 Photonics Res. 11 897
[20] Cheng L, Mao S, Li Z, Han Y and Fu H 2020 Micromachines 11 666
[21] Xie Y, Nie M and Huang S W 2024 Appl. Phys. Lett. 124 051108
[22] Ma W, Liu Z, Kudyshev Z A, Boltasseva A, Cai W and Liu Y 2021 Nat. Photonics 15 77
[23] Mourgias-Alexandris G, Moralis-Pegios M, Tsakyridis A, et al. 2022 Nat. Commun. 13 5572
[24] Hong L, Liu Y, Xu M and Deng W 2025 Chin. Phys. B 34 018705
[25] Cui J, Li Y, Zhao C and Zheng W 2023 Chin. Phys. B 32 096101
[26] Tu X, Xie W, Chen Z, Ge M F, Huang T, Song C and Fu H 2021 J. Light. Technol. 39 2790
[27] Adibnia E, Ghadrdan M and Mansouri-Birjandi M A 2025 Opt. Laser Technol. 181 111766
[28] Hooten S, Beausoleil R G and Van Vaerenbergh T 2021 Nanophotonics 10 3843
[29] Mengu D, Sakib RahmanMS, Luo Y, Li J, Kulce O and Ozcan A 2022 Adv. Opt. Photonics 14 209
[30] Li Y, Zhang Y, Wang Y, et al. 2024 Adv. Opt. Mater. 12 2302657
[31] Gostimirovic D andWinnie N Y 2018 IEEE J. Sel. Top. Quantum Electron. 25 1
[32] Zeng Z, Huang Q and He S 2023 Prog. Electromagn. Res. M 119 143
[33] Chen X, Lin D, Zhang T, Zhao Y, Liu H, Cui Y, Hou C, He J and Liang S 2023 Appl. Opt. 62 2924
[34] Kang S, Gao F, Yu X, Bo F, Zhang G and Xu J 2023 JOSA B 40 D21
[35] Augenstein Y, Repan T and Rockstuhl C 2023 ACS Photonics 10 1547
[36] Cheng L, Mao S, Tu X and Fu H Y 2021 J. Light. Technol. 39 5902
[37] Hammond A M, Slaby J B, ProbstMJ and Ralph S E 2022 Opt Express 30 31058
[38] Dai M, Ma L, Xu Y, Lu M, Liu X and Chen Y 2015 Opt Express 23 1691 |
| No Suggested Reading articles found! |
|
|
Viewed |
|
|
|
Full text
|
|
|
|
|
Abstract
|
|
|
|
|
Cited |
|
|
|
|
Altmetric
|
|
blogs
Facebook pages
Wikipedia page
Google+ users
|
Online attention
Altmetric calculates a score based on the online attention an article receives. Each coloured thread in the circle represents a different type of online attention. The number in the centre is the Altmetric score. Social media and mainstream news media are the main sources that calculate the score. Reference managers such as Mendeley are also tracked but do not contribute to the score. Older articles often score higher because they have had more time to get noticed. To account for this, Altmetric has included the context data for other articles of a similar age.
View more on Altmetrics
|
|
|
