Inverse design and experimental realization of plasma metamaterials
arXiv:2307.15221 · doi:10.1103/PhysRevApplied.20.044017
Abstract
We apply inverse design methods to produce two-dimensional triangular-lattice plasma metamaterial (PMM) devices which are then constructed and demonstrated experimentally. Finite difference frequency domain simulations are used along with forward-mode automatic differentiation to optimize the plasma densities of each of the plasma elements in the PMM to perform beam steering and demultiplexing under transverse magnetic polarization. The optimal device parameters are then used to assign plasma density values to elements that make up an experimental version of the device. Device performance is evaluated against both the simulated results and human-designed alternatives, showing the benefits and disadvantages of in-silico inverse design and paving the way for future fully in-situ optimization.
22 pages, 8 figures
References in corpus (7)
- Deep neural networks for the evaluation and design of photonic devices
- Experimentally realized in situ backpropagation for deep learning in nanophotonic neural networks
- Inverse design in photonics by topology optimization: tutorial
- Inverse design of photonic crystals through automatic differentiation
- Fullwave Maxwell inverse design of axisymmetric, tunable, and multi-scale multi-wavelength metalenses
- Inverse design of plasma metamaterial devices for optical computing
- Inverse design of plasma metamaterial devices with realistic elements