| Metamaterials are artificial structural media composed by sub-wavelength electromagnetic units with periodic or specifically aperiodic spatial arrangement. By carefully choosing the unit type and proper arrangement, they may exhibit exotic properties not found in nature. Nowadays, metamaterials with compact size and novel electromagnetic response provide a new way for the development of terahertz functional devices, which greatly boosts the progress of the terahertz technology. Recently the use of optimization method, especially the newly developed covariance matrix adaptation evolutionary strategy (CMA-ES), in the design of terahertz metamaterial devices has aroused great interests within the scientific community. CMA-ES is an improved alternative to well-established algorithms used in electromagnetic optimization. Compared with the classical optimization strategies, CMA-ES finds acceptable solutions more efficiently and more reliably. This thesis centres on the efficient design of terahertz functional devices based on metamaterials based on CMA-ES. The main contents of this thesis are as follows:(1) The principle and implementation of the CMA-ES have been discussed in detail. Two fitness functions are solved to find out the advantages of the CMA-ES.(2) One single-layered and double-layered terahertz metamaterial absorbers are designed using CMA-ES for large bandwidth and high absorption ratio. In the meanwhile, the design of a three-layered broadband terahertz absorber and one ultrabroadband plasmonic absorber are also optimized compared with those published results.(3) A terahertz low-scattering metasurface is designed using CMA-ES for best array configuration. Altogether eight windmill-like elements are used in the optimization, and excellent RCS reduction performance is demonstrated within a broad spectral range. |
