2026-10-13T12:01:08Z https://www.scidb.cn/oai

10.57760/sciencedb.j00186.00979 2026-05-06T13:44:22Z

2026-05-06 onperiodic Dielectric Laser Accelerator Optimization Dataset doi:10.57760/sciencedb.j00186.00979 en This dataset presents structured parameter–performance mappings for the optimization of nonperiodic dielectric laser accelerator (DLA) structures, generated through numerical simulations guided by a deep learning–based design framework. It includes multi-dimensional geometric parameters of dielectric grating units—such as pillar length, width, and rotation angle—together with physical conditions including optical field intensity, particle injection energy, and incidence angles, all organized in tabular form with weak inter-feature correlations. Each data entry corresponds to a unique structural configuration and its resulting particle energy gain, enabling the modeling of complex relationships between design variables and accelerator performance. The dataset focuses on nonperiodic structures that overcome the limitations of conventional periodic designs by enabling sustained acceleration and improved energy transfer efficiency. It is suitable for machine learning–based modeling, surrogate modeling, inverse design, and data-driven optimization, and provides a valuable resource for research in accelerator physics, photonics, and intelligent design of compact high-gradient DLA systems. Dielectric Laser Accelerator; optimization; dataset wang du PUBLIC https://creativecommons.org/licenses/by/4.0/ dataset Science Data Bank