基于遗传算法的超材料吸收器逆向设计研究

Research on Inverse Design of Metamaterial Absorbers Based on Genetic Algorithm

通过MATLAB编码图案控制Computer Simulation Technology(CST)Microwave Studio计算模拟软件,针对电磁超材料吸收器正向设计只能基于规则图案的现状,采用遗传算法对超材料吸收器进行逆向设计研究.以自动仿真为基础,用CST计算随机不同的初始图形结果,然后在遗传算法中迭代优化,不断用CST计算新出现图形的吸收率,最后优化出吸收率高的吸收器.首先进行单目标遗传算法优化设计吸收器,结果表明,在各个太赫兹波段,最高吸收峰的吸收率均大于96%,次高峰大于94%,平均吸收率均大于87%.然后进行多目标遗传算法优化设计,不再局限于一个单独的吸收率指标,能够加上如非目标波段的低吸收率的性能指标,实现单频吸收的功能.

In the paper,we present the research on inverse design of metamaterials absorbers based on the genetic algorithm,through controlling CST Microwave Studio computational simulation software by the MATLAB coded pattern,which can break the restrictions of the regular patterns in the traditional absorbers.Firstly,based on automatic simulation with beginning of the different random initial patterns,calculating the absorption rate,and selecting the absorber with the highest absorptivity,single objective genetic algorithm optimization for designing absorbers are performed.The calculation results show that the average absorption rates are above 87%at every terahertz wave region,while the highest absorption peak values are greater than 96%,and the second highest peaks are greater than 94%.We also perform the multi-objective genetic algorithm optimization inverse designing,and the single-frequency absorber is achieved.