基于遗传算法的五模材料分层优化*

由于超材料的微结构尺度不能忽略，利用超材料微结构所集成的器件等效参数不是连续分布的，而是呈离散分布的。因此在利用超材料设计声学器件的过程中，需要对器件进行分层离散化，而分层方案的选取往往会影响器件的工作效果。该文基于五模材料水下隐身衣的设计理论，提出了一种基于遗传算法的五模材料隐身衣分层优化策略，并分别对优化前和优化后的分层隐身衣做了声学仿真。仿真结果表明采用优化策略后，隐身衣背景声场的散射能量和反射能量均有大幅度降低，并且对窄带探测信号和宽带探测信号均有较好的隐身效果，因此该分层优化策略能够显著提升隐身衣的隐身效果。该文所提出的分层优化策略为提升五模材料隐身衣性能提供了理论支持，并可以进一步推广至其他超材料器件的微结构设计中。

A layered optimization strategy of pentamode materials based on genetic algorithm

Considering that the microstructure scale of metamaterials can not be ignored, the equivalent parameters of acoustic devices made by metamaterials are not continuous but discrete. In the process of design acoustic devices based on the metamaterials, it is necessary to layer the devices, and the selection of layering scheme often affects the working effect of the devices. In this paper, an optimization strategy used for layering the pentamode material cloak is proposed based on the genetic algorithm. The acoustic simulations of the cloak before and after optimization are carried out respectively, which shows that the scattered and reflected acoustic energy of the background acoustic field are greatly reduced after optimization for both narrowband and broadband detection signals. The proposed optimization strategy has a good performance, which can be extended to other types of metamaterial structure design.