Inverse design of photonic nanostructures based on machine learning (ML) methods has recently attracted much attention. An artificial neural network (ANN) showcases good performance on the photonic inverse design problem. However, there are still many unsolved issues regarding an efficient way to find a geometry that yields the target response by data-driven ML approaches. The design of air mass (AM) 1.5G filters for solar simulators represents such a challenging case. Here, we propose and show that a recurrent neural adjoint method is efficient in optimizing a multilayer optical filter for that purpose. Two examples of inverse design and optimization for an AM 1.5G filter with ${{\rm Si}_{3}} {{\rm N}_{4}} / {{\rm SiO}_{2}}$ and (${{\rm Si}_{3}} {{\rm N}_{4}} / {{\rm SiO}_{2}}$)/(${{\rm Ta}_{2}} {{\rm O}_{5}} / {{\rm SiO}_{2}}$) films at a different spectrum band (e.g., $\lambda = 280\;{\rm nm} - 800\;{\rm nm}$ and $\lambda = 280\;{\rm nm} - 1350\;{\rm nm}$) have been demonstrated. By comparing several strategies based on ANN approaches, a generic and efficient scheme is presented for photonic multilayer film structure engineering, which we believe could be applied to various photonic device designs.

中文翻译：

---

通过递归神经伴随方法逆向设计光学薄膜滤波器：以太阳模拟器为例

基于机器学习（ML）方法的光子纳米结构的逆向设计最近引起了很多关注。人工神经网络 (ANN) 在光子逆向设计问题上表现出良好的性能。然而，关于通过数据驱动的 ML 方法找到产生目标响应的几何的有效方法，仍然存在许多未解决的问题。用于太阳能模拟器的空气质量 (AM) 1.5G 过滤器的设计代表了这样一个具有挑战性的案例。在这里，我们提出并表明循环神经伴随方法在为此目的优化多层滤光片方面是有效的。具有${{\rm Si}_{3}} {{\rm N}_{4}} / {{\rm SiO}_{2}}$的 AM 1.5G 滤波器逆向设计和优化的两个示例和 ( ${{\rm Si}_{3}} {{\rm N}_{4}} / {{\rm SiO}_{2}}$ )/(${{\rm Ta}_{2}} {{\rm O}_{5}} / {{\rm SiO}_{2}}$ ) 薄膜在不同的光谱带（例如，$\lambda = 280\;{\rm nm} - 800\;{\rm nm}$和$\lambda = 280\;{\rm nm} - 1350\;{\rm nm}$ ) 已被证明。通过比较基于 ANN 方法的几种策略，提出了一种通用且有效的光子多层膜结构工程方案，我们相信该方案可以应用于各种光子器件设计。