Pilozzi Laura, Farrelly Francis A, Marcucci Giulia, Conti Claudio
Institute for Complex Systems, National Research Council (ISC-CNR), Via dei Taurini 19, 00185 Rome, Italy.
Department of Physics, University Sapienza, Piazzale Aldo Moro 5, 00185 Rome, Italy.
Nanotechnology. 2021 Apr 2;32(14):142001. doi: 10.1088/1361-6528/abd508.
We propose the use of artificial neural networks to design and characterize photonic topological insulators. As a hallmark, the band structures of these systems show the key feature of the emergence of edge states, with energies lying within the energy gap of the bulk materials and localized at the boundary between regions of distinct topological invariants. We consider different structures such as one-dimensional photonic crystals, [Formula: see text]-symmetric chains and cylindrical systems and show how, through a machine learning application, one can identify the parameters of a complex topological insulator to obtain protected edge states at target frequencies. We show how artificial neural networks can be used to solve the long-standing quest for a solution to inverse problems solution and apply this to the cutting edge topic of topological nanophotonics.
我们提议使用人工神经网络来设计和表征光子拓扑绝缘体。作为一个标志，这些系统的能带结构展现出边缘态出现的关键特征，其能量位于体材料的能隙内，并局域在具有不同拓扑不变量的区域之间的边界处。我们考虑了不同的结构，如一维光子晶体、[公式：见原文]对称链和圆柱系统，并展示了如何通过机器学习应用来识别复杂拓扑绝缘体的参数，以在目标频率处获得受保护的边缘态。我们展示了人工神经网络如何可用于解决长期以来对逆问题解决方案的探索，并将其应用于拓扑纳米光子学这一前沿主题。
