CLD-Based Automatic Frequency Selective Surface (FSS) Inverse Design

This work presents a deep learning-based method for the inverse design of stopband frequency selective surfaces (FSSs) with the CNN-LSTM-DNN (CLD) model. With the designed model, we can directly obtain the structure geometric parameters by inputting the CLD model with stopband FSS's S-parameters. Two cases of Jerusalem stopband FSS validate the proposed method can quickly fit the requirement.