Design of Wideband Frequency Selective Surface Based on the Combination of the Equivalent Circuit Model and Deep Learning

Abstract

In this letter, a design method for frequency selective surface (FSS) based on the equivalent circuit model (ECM) and deep learning is proposed. In the proposed method, the ECM depicting the target transmission characteristic is first established. Then, with the aid of the established ECM, the target transmission characteristic is converted into the desired transmission curve and the initial topology of the FSS structure is determined. The cascade network is adopted to realize the rapid inverse design of the FSS structure. Benefiting from the established ECM, the high quality of the training data and proper description of the design target can be ensured in deep learning. For verification, an ultra-wideband FSS structure with a passband ranging from 3.1 to 12.05GHz and a stopband ranging from 12.85 to 20GHz is designed. Both simulated and experimental results show that the designed FSS provides an ultra-wide passband with flat-top and fast roll-off characteristics, covering 98.9% target passband and 95.3% target stopband.