Sample-efficient inverse design of freeform nanophotonic devices with physics-informed reinforcement learning

to related objects

Abstract: Finding an optimal device structure in the vast combinatorial design space of freeform nanophotonic design has been an enormous challenge. In this study, we propose physics-informed reinforcement learning (PIRL) that combines the adjoint-based method with reinforcement learning to improve the sample efficiency by an order of magnitude compared to conventional reinforcement learning and overcome the issue of local minima. To illustrate these advantages of PIRL over other conventional optimization algorithms, we design a family of one-dimensional metasurface beam deflectors using PIRL, exceeding most reported records. We also explore the transfer learning capability of PIRL that further improves sample efficiency and demonstrate how the minimum feature size of the design can be enforced in PIRL through reward engineering. With its high sample efficiency, robustness, and ability to seamlessly incorporate practical device design constraints, our method offers a promising approach to highly combinatorial freeform device optimization in various physical domains.

Location: Deutsche Nationalbibliothek Frankfurt am Main

Extent: Online-Ressource

Language: Englisch

Bibliographic citation: Sample-efficient inverse design of freeform nanophotonic devices with physics-informed reinforcement learning ; volume:13 ; number:8 ; year:2024 ; pages:1483-1492 ; extent:10
Nanophotonics ; 13, Heft 8 (2024), 1483-1492 (gesamt 10)

Creator: Park, Chaejin
Kim, Sanmun
Jung, Anthony W.
Park, Juho
Seo, Dongjin
Kim, Yongha
Park, Chanhyung
Park, Chan Y.
Jang, Min Seok

DOI: 10.1515/nanoph-2023-0852

URN: urn:nbn:de:101:1-2024040616000747418693

Rights: Open Access; Der Zugriff auf das Objekt ist unbeschränkt möglich.

Last update: 14.08.2025, 10:55 AM CEST

Data provider

This object is provided by:
Deutsche Nationalbibliothek. If you have any questions about the object, please contact the data provider.

Show original at data provider

Associated

Park, Chaejin
Kim, Sanmun
Jung, Anthony W.
Park, Juho
Seo, Dongjin
Kim, Yongha
Park, Chanhyung
Park, Chan Y.
Jang, Min Seok

Other Objects (12)

Inverse design of free-form nanophotonic devices

Physics-Informed Particle-Based Reinforcement Learning for Autonomy in Signalized Intersections

Stiff-PDEs and Physics-Informed Neural Networks

Physics-informed shape optimization using coordinate projection

Physics‐Informed Inverse Design of Programmable Metasurfaces

Free-form optimization of nanophotonic devices: from classical methods to deep learning

Physics-informed regression of implicitly-constrained robot dynamics

Predicting glass structure by physics-informed machine learning

Physics-Informed Masked Autoencoder for active sparse imaging

Physics-Informed Deep Learning for Advanced Medical Ultrasound

A physically-informed machine learning model for freeform bending

Learning free-surface flow with physics-informed neural networks

Inverse design of free-form nanophotonic devices

Physics-Informed Particle-Based Reinforcement Learning for Autonomy in Signalized Intersections

Stiff-PDEs and Physics-Informed Neural Networks

Physics-informed shape optimization using coordinate projection

Physics‐Informed Inverse Design of Programmable Metasurfaces

Free-form optimization of nanophotonic devices: from classical methods to deep learning

Physics-informed regression of implicitly-constrained robot dynamics

Predicting glass structure by physics-informed machine learning

Physics-Informed Masked Autoencoder for active sparse imaging

Physics-Informed Deep Learning for Advanced Medical Ultrasound

A physically-informed machine learning model for freeform bending

Learning free-surface flow with physics-informed neural networks

Inverse design of free-form nanophotonic devices

Physics-Informed Particle-Based Reinforcement Learning for Autonomy in Signalized Intersections

Stiff-PDEs and Physics-Informed Neural Networks

Physics-informed shape optimization using coordinate projection

Physics‐Informed Inverse Design of Programmable Metasurfaces

Free-form optimization of nanophotonic devices: from classical methods to deep learning

Physics-informed regression of implicitly-constrained robot dynamics

Predicting glass structure by physics-informed machine learning

Physics-Informed Masked Autoencoder for active sparse imaging

Physics-Informed Deep Learning for Advanced Medical Ultrasound

A physically-informed machine learning model for freeform bending

Learning free-surface flow with physics-informed neural networks

Cultural heritage institutions wishing to register will find more information here.

Fields marked * need to be filled in.

Username*

Please enter your username

Email*

Please enter your email address

Please do not fill this field

First name

Last name

Password*

Please enter your password

Confirm password*

Please enter the same password

I have read the terms of use and the privacy policy for the collection of personal data and accept them. *

This field is required.

I would like to subscribe to the newsletter of the Deutsche Digitale Bibliothek. See newsletter subscription info.

Account created

Your "My DDB" account has been successfully created. Before you can log in to your account, you must click the confirmation link in the message we just sent to the email address you provided.

Sample-efficient inverse design of freeform nanophotonic devices with physics-informed reinforcement learning

Object Details

References and Relationships

Contributors, Places and Time

Further information

Data provider

Associated

Other Objects (12)

Inverse design of free-form nanophotonic devices

Physics-Informed Particle-Based Reinforcement Learning for Autonomy in Signalized Intersections

Stiff-PDEs and Physics-Informed Neural Networks

Physics-informed shape optimization using coordinate projection

Physics‐Informed Inverse Design of Programmable Metasurfaces

Free-form optimization of nanophotonic devices: from classical methods to deep learning

Physics-informed regression of implicitly-constrained robot dynamics

Predicting glass structure by physics-informed machine learning

Physics-Informed Masked Autoencoder for active sparse imaging

Physics-Informed Deep Learning for Advanced Medical Ultrasound

A physically-informed machine learning model for freeform bending

Learning free-surface flow with physics-informed neural networks

Inverse design of free-form nanophotonic devices

Physics-Informed Particle-Based Reinforcement Learning for Autonomy in Signalized Intersections

Stiff-PDEs and Physics-Informed Neural Networks

Physics-informed shape optimization using coordinate projection

Physics‐Informed Inverse Design of Programmable Metasurfaces

Free-form optimization of nanophotonic devices: from classical methods to deep learning

Physics-informed regression of implicitly-constrained robot dynamics

Predicting glass structure by physics-informed machine learning

Physics-Informed Masked Autoencoder for active sparse imaging

Physics-Informed Deep Learning for Advanced Medical Ultrasound

A physically-informed machine learning model for freeform bending

Learning free-surface flow with physics-informed neural networks

Inverse design of free-form nanophotonic devices

Physics-Informed Particle-Based Reinforcement Learning for Autonomy in Signalized Intersections

Stiff-PDEs and Physics-Informed Neural Networks

Physics-informed shape optimization using coordinate projection

Physics‐Informed Inverse Design of Programmable Metasurfaces

Free-form optimization of nanophotonic devices: from classical methods to deep learning

Physics-informed regression of implicitly-constrained robot dynamics

Predicting glass structure by physics-informed machine learning

Physics-Informed Masked Autoencoder for active sparse imaging

Physics-Informed Deep Learning for Advanced Medical Ultrasound

A physically-informed machine learning model for freeform bending

Learning free-surface flow with physics-informed neural networks

Related objects

Reset password