We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Free-form optimization of nanophotonic devices: from classical methods to deep learning.
- Authors
Park, Juho; Kim, Sanmun; Nam, Daniel Wontae; Chung, Haejun; Park, Chan Y.; Jang, Min Seok
- Abstract
Keywords: adjoint method; free-form optimization; machine learning; photonic device design; reinforcement learning EN adjoint method free-form optimization machine learning photonic device design reinforcement learning 1809 1845 37 05/12/22 20220401 NES 220401 1 Introduction Nanophotonic devices control the behavior of light on the subwavelength scale by harnessing the interaction of nanometer-scale objects with light. I A i is the space in which the action of an agent is defined, HT <math overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><mi>a</mi><mo> </mo><mi>A</mi></math> ht . The main difficulty is in evaluating HT <math overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><mrow><mrow><mo> </mo><mi mathvariant="bold">x</mi><mo>/</mo><mo> </mo><mi mathvariant="bold">p</mi></mrow></mrow></math> ht , which is typically implicit in the governing equation such as Eq. (1). (1) to I i B x b / I i B p b and HT <math overflow="scroll" xmlns="http://www.w3.org/1998/Math/MathML"><mrow><mrow><mo> </mo><mi mathvariant="bold">b</mi><mo>/</mo><mo> </mo><mi mathvariant="bold">p</mi></mrow><mo>-</mo><mrow><mo>(</mo><mrow><mo> </mo><mi>A</mi><mo>/</mo><mo> </mo><mi mathvariant="bold">p</mi></mrow><mo>)</mo></mrow><mi mathvariant="bold">x</mi></mrow></math> ht , respectively. The two schemes approach a faithful representation of the whole design space as their number of parameters increases, i.e. the number of grids in the grid representation and the number of spectral basis vectors in the Fourier representation.
- Subjects
DEEP learning; PLASMONICS; CONJUGATE gradient methods; ORGANIC light emitting diode efficiency; REINFORCEMENT learning
- Publication
Nanophotonics (21928606), 2022, Vol 11, Issue 9, p1809
- ISSN
2192-8606
- Publication type
Article
- DOI
10.1515/nanoph-2021-0713