Reinforcement Learning and Differentiable Simulations for Autonomous Tuning and Control of Linear Particle Accelerators

Kaiser, Jan

doi:10.15480/882.16060

Book/Dissertation / PhD Thesis

PUBDB-2026-00725

Reinforcement Learning and Differentiable Simulations for Autonomous Tuning and Control of Linear Particle Accelerators

Kaiser, J. (Corresponding author)DESY*

2025
Dr. Hut Verlag
ISBN: 978-3-8439-5683-3

Dr. Hut Verlag, Regelungstechnik 231 pages (2025) [10.15480/882.16060] = Dissertation, Hamburg University of Technology, 2025

This record in other databases:

Please use a persistent id in citations: doi:10.15480/882.16060

Abstract: Particle accelerators are sophisticated scientific facilities that require precise but time-consuming optimisation to achieve optimal performance. Considering benchmark tasks at the ARES and LCLS facilities, this dissertation proposes methods to deploy simulation-trained reinforcement learning (RL) policies for accelerator tuning zero-shot to the real world and novel tuning tasks, while comparing their performance to traditional methods. A high-speed differentiable beam dynamics simulator is developed to make collecting large datasets for RL feasible, and to enable a multitude of novel gradient-based accelerator applications. These contributions lay the groundwork for faster accelerator tuning to better working points, and enable new scientific discoveries.

Keyword(s): reinforcement learning ; differentiable simulation ; particle accelerators ; Natural Sciences and Mathematics::539: Matter; Molecular Physics; Atomic and Nuclear physics; Radiation; Quantum Physics ; Computer Science, Information and General Works::006: Special computer methods::006.3: Artificial Intelligence::006.31: Machine Learning ; Technology::681: Precision Instruments and Other Devices::681.2: Testing, Measuring, Sensing Instruments