Physical Information-Embedded Deep Learning for Forward Prediction and Inverse Design of Nanophotonic Devices

Nanophotonic devices have enabled electromagnetic manipulation with unprecedented accuracy and functionalities. With the increased functionality requirements of such devices and the progress of fabrication technologies, trends toward miniaturization and multi-functionality have emerged, making forwa...

Full description

Saved in:
Bibliographic Details
Published in:Journal of lightwave technology 2021-10, Vol.39 (20), p.6498-6508
Main Authors: Song, Yuchen, Wang, Danshi, Qin, Jun, Li, Jin, Ye, Han, Zhang, Zhiguo, Chen, Xue, Zhang, Min, Boucouvalas, Anthony C.
Format: Article
Language:English
Subjects:
Accuracy
Broadband
Crosstalk
Data models
Deep learning
Demultiplexers
Embedding
Feature extraction
Finite element analysis
forward prediction
Inverse design
Iterative methods
Mathematical model
Miniaturization
Multidimensional methods
nanophotonic devices
Nanoscale devices
Optimization
Power splitters
Prediction models
Predictions
Predictive models
Topology
Training
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Nanophotonic devices have enabled electromagnetic manipulation with unprecedented accuracy and functionalities. With the increased functionality requirements of such devices and the progress of fabrication technologies, trends toward miniaturization and multi-functionality have emerged, making forward response prediction and inverse design increasingly complicated. In this paper, a physical information-embedded deep learning model for flexible forward prediction and multifunctional inverse design is proposed. Besides the overwhelming advantage of computing efficiency over conventional iterative optimization methods, the proposed method exhibits strong generalization on different wavelengths and polarizations by embedding the corresponding physical information into the model through effective sparse parameter method (SPM). For the forward prediction, responses can be predicted among the 1300-1550 nm wavelength band and two polarizations (TM/TE) with high accuracy. For the inverse design, with the assistance of forward prediction models and broadband training methods, multi-dimensional information for the target can be considered simultaneously to retrieve the satisfactory desired designs topologies in an unsupervised way. The multi-functionality of inverse design is demonstrated by implementing wavelength demultiplexers with small crosstalk and power splitters with arbitrary splitting ratios on a 250 nm broadband wavelength in a compact footprint of 2 × 2 μm 2 without additional heavy training of the model.
ISSN:0733-8724
1558-2213
DOI:10.1109/JLT.2021.3103758