In this paper, a compact branch microstrip coupler with the simple structure using T-shaped resonators and open-ended stubs at 1.8 GHz is designed and fabricated. The proposed coupler creates two transmission zeros at 3.4 and 3.87 GHz, with more than 50 dB attenuation level, which resulted in a good harmonic suppression at 2nd harmonic. Artificial neural network (ANN) has been utilized to extract the transfer function of the proposed coupler resonator for the first time, so the values of the transmission zeros can be located in the desired frequency. This technique can also be used for the other electrical devices. The designed device has a small size of 28 mm × 30 mm (0.9 λ × 0.95 λ), which shows more than 35% size reduction compared to the typical branch-line coupler with λ/4 branch lines. The proposed coupler is simulated using Advanced Design System (ADS) software and fabricated on Rogers Duroid 5880 substrate (ε r = 2.2, h = 31 mil). The simulation and measurement results verify the correct performance of the designed coupler.

中文翻译：

---

基于人工神经网络的结构简单的小型化支线微带耦合器设计

本文设计并制作了一种结构简单的紧凑型分支微带耦合器，该耦合器采用T形谐振器和1.8 GHz的开放式短截线。所提出的耦合器在 3.4 和 3.87 GHz 处产生两个传输零点，衰减水平超过 50 dB，从而在 2 次谐波处产生良好的谐波抑制。人工神经网络 (ANN) 首次被用于提取所提出的耦合器谐振器的传递函数，因此传输零点的值可以位于所需的频率中。这种技术也可以用于其他电气设备。所设计的器件尺寸为 28 mm × 30 mm (0.9λ× 0.95 λ），与具有 λ/4 支线的典型支线耦合器相比，尺寸减小了 35% 以上。所提出的耦合器使用高级设计系统 (ADS) 软件进行仿真，并在 Rogers Duroid 5880 基板上制造（ε r= 2.2,H= 31 百万）。仿真和测量结果验证了所设计耦合器的正确性能。