Optimization of the transfer line against collective effects using physics-constrained generative phase space reconstruction

MOCG004
22 Sept 2025, 14:30
15m
Grand Ballroom (Palmer House Hilton Chicago)

Grand Ballroom

Palmer House Hilton Chicago

17 East Monroe Street Chicago, IL 60603, United States of America
Contributed Oral Presentation MC13: Artificial Intelligence & Machine Learning MOCG MC13 Artificial Intelligence and Machine Learning

Speaker

Seongyeol Kim (Pohang Accelerator Laboratory)

Description

Optimization of the transfer line against collective effects such as space charge and coherent synchrotron radiation (CSR) effects is crucial to preserve the beam quality. While simple conventional diagnostic methods provide ensemble averaged beam parameters or limited information of phase space, they are still limited in obtaining precise, complete 6-dimensional phase space with all the correlations due to hardware and dedicated time requirements. A generative phase space reconstruction method (GPSR) has been developed as a robust diagnostic framework that reconstructs complete 6-dimensional phase space. Here we show a physics-constrained GPSR model that incorporates known physical parameters, such as RMS beam sizes and emittances, as constraints. We performed simulated demonstrations at the Pohang Accelerator Laboratory X-ray Free Electron Laser facility that the Physics-informed GPSR can be performed for complete 6-dimensional phase space. Furthermore, by using the reconstructed phase space, we performed non-differentiable particle tracking simulations to investigate the phase space evolution against the space charge and CSR along the bunch compressor. We present the trend of predicted CSR-induced emittance growth, which closely matches the ground truth.

Funding Agency

This work was supported by the National Research Foundation of Korea (NRF) grant (RS-2024-00347026), funded by the Korea government (MSIT).

Author

Seongyeol Kim (Pohang Accelerator Laboratory)

Co-authors

Ryan Roussel (SLAC National Accelerator Laboratory) Juan Pablo Gonzalez-Aguilera (University of Chicago) Mr Kim Gyujin (Pohang Accelerator Laboratory) Auralee Edelen (SLAC National Accelerator Laboratory) Haeryong Yang (Pohang Accelerator Laboratory)

Presentation materials

There are no materials yet.