It is of great significance to develop and design terahertz metamaterials with superior sensing characteristics, since such metameterials can be widely used in the sensing fields of biology, medicine and so on. Traditional metamaterial design is mainly based on impedance matching and equivalent circuit theory. It is designed through model design, parameter scanning and manual optimization. It needs hundreds of parameter scanning operations and consumes a lot of computer resources. This design method has high requirements for the designers’ electromagnetic theory and simulation experience, and it depends heavily on the designers’ metamaterial design experience and intuition. In recent years, the metamaterial inverse design method based on artificial intelligence technology has developed rapidly. However, it is difficult for this method to deal with the design of different sizes, the optimization is easy to fall into the local optimal state, and there is no special design method for metamaterial absorber. To solve the above problems, a rapid design method of metamaterial absorber based on improved particle swarm optimization algorithm is proposed in this paper. The absorption characteristics of metamaterial absorber are automatically calculated by using automatic peak seeking technology, so as to realize the optimization and inverse design of metamaterial metal resonance layer.

In this paper, the metamaterial metal resonance layer is expressed as a matrix, and the key parameters of the absorptivity curve of metamaterial are read by using the automatic peak seeking algorithm. The rapid design of metamaterial absorber is realized based on CST-MATLAB joint simulation technology and particle swarm optimization algorithm with linearly decreasing weight. During the use of the software, the designer inputs the design parameters such as the period, thickness and dielectric constant of metamaterial structural unit, and uses the particle swarm optimization algorithm to automatically carry out optimization iterative calculation, so as to output the qualified "optimal" metamaterial absorber design results. In order to further improve the sensing performance of metamaterials, taking the metamaterial structure designed by the method proposed in this paper as an example, a new scheme of metamaterial absorber based on microcavity structure is proposed, and in-depth simulation research is carried out.

Compared with other design methods, the metamaterial rapid inverse design method proposed in this paper saves a lot of time and reduces manual participation (Table 1). The design results show that using the metamaterial rapid design method based on particle swarm optimization algorithm can design results similar to manual design (Fig. 4), and achieve almost perfect absorption of incident electromagnetic wave at 0.3122 THz (Fig. 5). It has high sensing performance (Fig. 6). By replacing the original intermediate dielectric layer with the analyte to be measured, the coupling between the analyte to be measured and the resonant magnetic field can be effectively strengthened, and the sensing characteristics of the metamaterial absorber are improved (Table 2).

In this paper, a rapid inverse design method of metamaterial based on particle swarm optimization and automatic peak seeking algorithm is proposed. Taking the metamaterial structure designed by the method proposed in this paper as an example, the design scheme of metamaterial absorber based on microcavity structure is proposed, and in-depth simulation research is carried out. The results show that the rapid design method can not only save the design time of metamaterial metal resonance layer and simplify the design process, but also design a metamaterial structure similar to manual design results. Compared with the traditional design method, the inverse design method of metamaterial absorber based on particle swarm optimization avoids the requirements of designers’ knowledge and experience. The metamaterial rapid design method studied in this paper provides a technical basis for the subsequent application of artificial intelligence technology in metamaterial design, and has guiding significance for the mechanism analysis and sensor design of metamaterial absorber in the future. Compared with the traditional structure metamaterials, although the absorptivity, resonant frequency, full-width at half-maximum and several other parameters are somewhat worse, the microcavity metamaterials have better sensing characteristics, simplify the metamaterial design process, and better meet the sensor design requirements. The metamaterial structure of microcavity structure provides a new solution to the application design of metamaterial absorber.