结合Skyrme模型和深度学习研究原子核的结团结构

围绕原子核结团结构的研究热点，基于π介子场的非线性理论Skyrme模型真实而又生动地呈现了“拓扑”与核物理交叉融合在原子核结团结构研究中的独特优势。因此，发展Skyrme模型，精确描述并预测原子核结团结构，探索结团形成机制尤为重要。本项目拟基于前期工作，在Skyrme框架下尝试结合深度学习在获取内部规律、优化研究目标、内插外推特性方面的优势，通过已知原子核的结团信息对深度学习进行训练，优化Skyrme模型对原子核结团结构理论描述，并探索结团结构形成的物理机制。同时通过深度学习预测未知原子核的结团结构。最后，以Skyrme模型与宏观-微观质量模型的类液滴思想共性为切入点，结合前期关于库仑能项的研究，提取并优化Skyrmion孤子解的结合能。项目旨在实现Skyrmion对原子核结团结构和结合能的精确描述和预测，以及对结团形成机制的探索。

Focusing on the scientific frontier of the cluster structure in the nucleus, the Skyrme model based on the nonlinear theory of the π meson field truly and vividly presents the unique advantages of the cross-fusion of "topology" and nuclear physics in the study of the cluster structure of atomic nuclei in nuclear physics. Therefore, it is particularly important to develop the Skyrme model, accurately describe and predict the cluster structure of nuclei, and explore the mechanism of cluster formation. This project is planned to be based on the previous work. Under the framework of Skyrme, it is the first attempt to combine the advantages of deep learning to obtain internal laws, optimize research goals, and interpolation and extrapolation characteristics. With the training of deep learning by using the clustering information of known nuclei, to optimize the theoretical description of the nucleus cluster structure, and to explore the physical mechanism of the cluster structure. At the same time, deep learning is used to predict the cluster structure of unknown nuclei. Finally, taking the commonality of droplets between the Skyrme model and the macro-micro mass model as the starting point, combined with the previous research on the Coulomb energy term, the binding energy of the Skyrmion soliton solution is extracted and optimized. Through this project, Skyrmion can be used to the description and prediction of the cluster structure and binding energy.