A knowledge-inherited learning for intelligent metasurface design and assembly

作     者：Yuetian Jia Chao Qian Zhixiang Fan Tong Cai Er-Ping Li Hongsheng Chen Yuetian Jia;Chao Qian;Zhixiang Fan;Tong Cai;Er-Ping Li;Hongsheng Chen

作者机构：ZJU-UIUC InstituteInterdisciplinary Center for Quantum InformationState Key Laboratory of Extreme Photonics and InstrumentationZhejiang UniversityHangzhou 310027China ZJU-Hangzhou Global Science and Technology Innovation CenterKey Lab.of Advanced Micro/Nano Electronic Devices&Smart Systems of ZhejiangZhejiang UniversityHangzhou 310027China Jinhua Institute of Zhejiang UniversityZhejiang UniversityJinhua 321099China Air and Missile Defense CollegeAir Force Engineering UniversityXi'an 710051China Shaoxing Institute of Zhejiang Universityzhejiang UniversityShaoxing321000China 

出 版 物：《Light(Science & Applications)》 (光（科学与应用)（英文版）)

年 卷 期：2023年第12卷第4期

页      面：680-690页

核心收录：

学科分类：070207[理学-光学] 07[理学] 08[工学] 0803[工学-光学工程] 0702[理学-物理学] 

基　　金：sponsored by the Key Research and Development Program of the Ministry of Science and Technology under Grants No.2022YFA1404704,2022YFA1404902,2022YFA1405200 the National Natural Science Foundation of China(NNSFC)under Grants No.11961141010,No.61975176,the Top-Notch Young Talents Program of China and the Fundamental Research Funds for the Central Universities 

主　　题：breakthrough consuming shortcomings 

摘      要：Recent breakthroughs in deep learning have ushered in an essential tool for optics and photonics,recurring in various applications of material design,system optimization,and automation *** learning-enabled on-demand metasurface design has been the subject of extensive expansion,as it can alleviate the time-consuming,low-efficiency,and experience-orientated shortcomings in conventional numerical simulations and physics-based ***,collecting samples and training neural networks are fundamentally confined to predefined individual metamaterials and tend to fail for large problem *** by object-oriented C++programming,we propose a knowledge-inherited paradigm for multi-object and shape-unbound metasurface inverse *** inherited neural network carries knowledge from theparentmetasurface and then is freely assembled to construct theoffspringmetasurface;such a process is as simple as building a container-type *** benchmark the paradigm by the free design of aperiodic and periodic metasurfaces,with accuracies that reach 86.7%.Furthermore,we present an intelligent origami metasurface to facilitate compatible and lightweight satellite communication *** work opens up a new avenue for automatic metasurface design and leverages the assemblability to broaden the adaptability of intelligent metadevices.