26–31 Jul 2026
Luskin Conference Center, UCLA
US/Pacific timezone

High-dimensional phase space reconstruction using generative modeling of beam distributions

31 Jul 2026, 09:45
30m
Auditorium (CNSI)

Auditorium

CNSI

Invited plenary Invited Talks

Speaker

Juan Pablo Gonzalez-Aguilera (SLAC National Accelerator Laboratory)

Description

Advanced accelerator concepts rely on the precise measurement and control of the beam distribution. However, conventional methods used to characterize precise features of the beam distribution in six-dimensional phase space require many measurements taken over tens of hours of beam time. In recent years, generative phase space reconstruction (GPSR) has been shown to significantly reduce the number of measurements needed to reconstruct the beam distribution in four-, five-, and six-dimensional phase space by using a generative model of the beam combined with differentiable beam dynamics simulations. In this talk, we present an overview of the GPSR method, including its implementation at the Argonne Wakefield Accelerator and other facilities. Furthermore, we show recent developments in GPSR that enable its use in real time for online tuning of beam distributions in advanced accelerator applications.

Working group WG5

Authors

Juan Pablo Gonzalez-Aguilera (SLAC National Accelerator Laboratory) Ryan Roussel (SLAC National Accelerator Laboratory) Auralee Edelen (SLAC National Accelerator Laboratory) Dylan Kennedy (SLAC National Accelerator Laboratory) Yiheng Ye (Stanford University) Lewis Boulton (DESY) Seongyeol Kim

Presentation materials

There are no materials yet.