Physics-informed neural networks for inverse problems in nano-optics and metamaterials.

@article{Chen2020PhysicsinformedNN,
  title={Physics-informed neural networks for inverse problems in nano-optics and metamaterials.},
  author={Yuyao Chen and Lu Lu and George Em Karniadakis and Luca Dal Negro},
  journal={Optics express},
  year={2020},
  volume={28 8},
  pages={
          11618-11633
        }
}
In this paper, we employ the emerging paradigm of physics-informed neural networks (PINNs) for the solution of representative inverse scattering problems in photonic metamaterials and nano-optics technologies. In particular, we successfully apply mesh-free PINNs to the difficult task of retrieving the effective permittivity parameters of a number of finite-size scattering systems that involve many interacting nanostructures as well as multi-component nanoparticles. Our methodology is fully… 

Figures from this paper

Physics-informed neural network for inversely predicting effective electric permittivities of metamaterials
We apply a physics-informed neural network framework for inversely retrieving the effective material parameters of a two-dimensional metamaterial from its scattered field(s). We show that by
Deep Learning the Electromagnetic Properties of Metamaterials—A Comprehensive Review
Deep neural networks (DNNs) are empirically derived systems that have transformed traditional research methods, and are driving scientific discovery. Artificial electromagnetic materials
WaveY-Net: physics-augmented deep-learning for high-speed electromagnetic simulation and optimization
We introduce WaveY-Net, a hybrid data- and physics-augmented convolutional neural network that can predict electromagnetic field distributions with ultra fast speeds and high accuracy for entire
Hyper-parameter tuning of physics-informed neural networks: Application to Helmholtz problems
TLDR
A hyper-parameter tuning procedure via Gaussian processes-based Bayesian optimization is applied to Helmholtz problems for bounded domains and a thorough study is conducted, focusing on performance and the applicability and necessity of the method.
Physics-informed Neural Network for Nonlinear Dynamics in Fiber Optics
TLDR
The results report here show that the PINN is not only an effective partial differential equation solver, but also a prospective technique to advance the scientific computing and automatic modeling in fiber optics.
Physics-informed neural networks for imaging and parameter retrieval of photonic nanostructures from near-field data
Physics-informed neural networks for imaging and parameter retrieval of photonic nanostructures from near-field data Yuyao Chen1 and Luca Dal Negro1, 2, 3 1)Department of Electrical & Computer
Physics-informed neural networks for solving parametric magnetostatic problems
TLDR
The objective of this paper is to investigate the ability of physics-informed neural networks to learn the magnetic field response as a function of design parameters in the context of a two-dimensional (2-D) magnetostatic problem.
Solving the wave equation with physics-informed deep learning
We investigate the use of Physics-Informed Neural Networks (PINNs) for solving the wave equation. Whilst PINNs have been successfully applied across many physical systems, the wave equation presents
Physics-informed neural networks with hard constraints for inverse design
TLDR
This work proposes a new deep learning method—physics-informed neural networks with hard constraints (hPINNs)—for solving topology optimization and demonstrates the effectiveness of hPINN for a holography problem in optics and a fluid problem of Stokes flow.
Modified physics-informed neural network method based on the conservation law constraint and its prediction of optical solitons
Based on conservation laws as one of the important integrable properties of nonlinear physical models, we design a modified physics-informed neural network method based on the conservation law
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 71 REFERENCES
Past Achievements and Future Challenges in 3D Photonic Metamaterials
Photonic metamaterials are man-made structures composed of tailored micro- or nanostructured metallo-dielectric sub-wavelength building blocks that are densely packed into an effective material. This
Past achievements and future challenges in the development of three-dimensional photonic metamaterials
Photonic metamaterials are man-made structures composed of tailored micro- or nanostructured metallodielectric subwavelength building blocks. This deceptively simple yet powerful concept allows the
Embedding Deep Learning in Inverse Scattering Problems
TLDR
A novel convolutional neural network architecture is proposed, termed the contrast source network, that learns the noise space components of the radiation operator that helps in producing high resolution solutions without any significant increase in computational costs.
Inverse design in nanophotonics
TLDR
A cross-section of key developments in this emerging field of photonic optimization is outlined: moving from a recap of foundational results to motivation of applications in nonlinear, topological, near-field and on-chip optics.
Plasmonic cloaking of cylinders: Finite length, oblique illumination and cross-polarization coupling
Metamaterial cloaking has been proposed and studied in recent years by following several interesting approaches. One of them, the scattering-cancelation technique or plasmonic cloaking, exploits the
Efficient and accurate inversion of multiple scattering with deep learning
TLDR
This paper designs and training a deep convolutional neural network that can invert multiple scattered measurements to produce a high-quality image of the refractive index and shows that the proposed approach is substantially faster and achieves higher imaging quality compared to the state-of-the-art methods based on optimization.
Invisibility and Cloaking: Origins, Present, and Future Perspectives
The development of metamaterials, i.e., artificially structured materials that interact with waves in unconventional ways, has revolutionized our ability to manipulate the propagation of
Effective medium theory for two-dimensional non-magnetic metamaterial lattices up to quadrupole expansions
We present a formulation for deriving effective medium properties of infinitely periodic two-dimensional metamaterial lattice structures beyond the conventional static and quasi-static limits. We
REVIEW ARTICLE: Plasmonic and metamaterial cloaking: physical mechanisms and potentials
TLDR
Some of the different cloaking mechanisms recently proposed in the literature, such as plasmonic cloaking based on scattering cancelation, coordinate-transformation cloaking and anomalous localized resonances for cloaking are overviewed, with particular attention to the physical phenomena involved.
...
1
2
3
4
5
...