Inverse design of metasurface optical filters using deep neural network with high degrees of freedom
Han Xiao, Ziyang Fan, Zeyang Liu, Chao Li, L. Jay Guo
Abstract
Abstract In order to obtain a metasurface structure capable of filtering light of a specific wavelength range in the visible band, the traditional methods usually traverse the space consisting of possible designs, searching for a potentially satisfactory structure by performing iterative calculations to solve Maxwell's equations. In this article, we propose a systematic method based on neural networks that can complete an inverse design process to solve the problem. Compared with the traditional methods, our method is much faster while competent to encompass a high degree of freedom to generate device structures, which can ensure that the spectra of generated structures resemble the desired ones. image