Deep learning-assisted design for metasurfaces with purified single resonant peak

Zhenxiang Shi; Haiou Lu; Bing Yan; Xinyu Yu; Min Zhang; Kai Ni; Qian Zhou; Xiaohao Wang

doi:10.1117/12.3006742

21 December 2023 Deep learning-assisted design for metasurfaces with purified single resonant peak

Zhenxiang Shi, Haiou Lu, Bing Yan, Xinyu Yu, Min Zhang, Kai Ni, Qian Zhou, Xiaohao Wang

Proceedings Volume 12966, AOPC 2023: AI in Optics and Photonics ; 129661C (2023) https://doi.org/10.1117/12.3006742
Event: Applied Optics and Photonics China 2023 (AOPC2023), 2023, Beijing, China

Abstract

Dielectric metasurfaces designed based on bound states in the continuum (BICs) exhibit extremely high quality factors, high sensitivity, and low loss, making them suitable for surface-enhanced infrared absorption and refractive index sensing. It has been demonstrated in some literatures that the symmetric property of the elliptical unit cell of the BICs mode can be broken by increasing the rotation angle, thereby achieving high-quality factor resonant structures. By varying the rotation angle, major and minor axis lengths of the ellipse, and the period of the bilayer elliptical unit, unique resonant properties can be achieved even without considering the existence of BIC. In this work, we have designed a deep learning-based transmission spectrum prediction network by combining the parameters of the elliptical unit. This network can replace traditional electromagnetic simulation calculations to quickly obtain the transmission spectrum of the target structure. While simulating 3000 sets of elliptical unit using the finite-difference time-domain method on an i7- 12700H processor requires 14 hours, using our neural network yields a transmission spectrum with prediction accuracy better than 10^-3 in less than 6.9 seconds, significantly improving the design efficiency. The network takes the image of metasurface unit as input data and couples the scaling factor that affects the period of the metasurface into the image data, making it possible to train and predict the spectra with different structures and ratios, effectively improving the generalization ability of the network.

(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Zhenxiang Shi, Haiou Lu, Bing Yan, Xinyu Yu, Min Zhang, Kai Ni, Qian Zhou, and Xiaohao Wang "Deep learning-assisted design for metasurfaces with purified single resonant peak", Proc. SPIE 12966, AOPC 2023: AI in Optics and Photonics , 129661C (21 December 2023); https://doi.org/10.1117/12.3006742

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
5 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Design and modelling

Electromagnetic simulation

Dielectrics

Neural networks

Refractive index

Finite-difference time-domain method

Mathematical optimization

Keywords/Phrases

Search In:

Publication Years