为了缅怀物理教育家赵凯华先生注重培养科学素质的教育理念，且贯彻人工智能赋能物理教学的政策方针，同时激发本科学生对前沿物理课题的研究兴趣，本文在理论和仿真中探讨拓扑层结构的非线性特征以及其作为非线性激活函数在光神经网络中的交叉应用.研究内容可作为物理系本科生专业前沿课程的开放性研究课题，帮助学生理解前沿学科拓扑光子学和光神经网络，有效推进教学与科研的结合.

In order to motivate and inspire the research interests of undergraduate students and help them to learn more about the frontier research area, we provide an example on how to combine the photonic neural network into the frontier research field of physics, specifically the layered structure in topological photonics. The results can be used as an open project for the frontier research courses for physics major senior undergraduate student. It can help them to deeply understand the basic models in topological photonics and artificial intelligent modeling method such as neural network. This paper provides a road towards developing the fusion between research and teaching.