In this study, a bioinspired computation (BIC) optimisation technique to synthesise the geometric dimensions of a lozenge-shaped broadband frequency selective surface (FSS) for electromagnetic wave filtering in satellite communications systems, for the X (8–12 GHz) and Ku (12–18 GHz) bands, is presented. A hybrid BIC technique, that combines the general regression neural network with a multi-objective flower pollination algorithm, is proposed. The objectives of the optimisation process consist of tuning the resonant frequency of the lozenge-shaped FSS to 15.2 GHz and operating bandwidth to 8 GHz, characterising a broadband framework. The cutoff frequencies to control the operating frequency range of the FSS, referring to transmission coefficient in decibels (dB), were obtained at a threshold of −10 dB. The proposed lozenge-shaped FSS, after optimisation of the geometrical dimensions of its unit cell, was fabricated on fiberglass substrate (FR4). The simulated and measured results show a good agreement.

中文翻译：

---

通过快速，准确的混合生物启发多目标技术优化的紧凑型超宽带FSS

在这项研究中，采用了一种生物启发式计算（BIC）优化技术来合成菱形形状的宽带频率选择性表面（FSS）的几何尺寸，用于卫星通信系统中的X（8–12 GHz）和Ku（介绍了12–18 GHz）频段。提出了一种将通用回归神经网络与多目标花授粉算法相结合的混合BIC技术。优化过程的目标包括将菱形FSS的谐振频率调整为15.2 GHz，将工作带宽调整为8 GHz，这是宽带框架的特征。在-10 dB的阈值下，获得了控制FSS的工作频率范围的截止频率（参考以分贝（dB）为单位的传输系数）。建议的菱形FSS，在优化其晶胞的几何尺寸之后，在玻璃纤维基板（FR4）上进行制造。仿真结果与实测结果吻合良好。