Learning-based Optimisation of Particle Accelerators Under Partial Observability Without Real-World Training

Kaiser, Jan; Eichler, Annika; Stein, Oliver

Contribution to a conference proceedings/Contribution to a book

PUBDB-2022-02795

Learning-based Optimisation of Particle Accelerators Under Partial Observability Without Real-World Training

Kaiser, J. (Corresponding author)Extern*DESY* ; Stein, O.DESY* ; Eichler, A.XFEL.EU*DESY*

2022
PMLR

Proceedings of the 39th International Conference on Machine Learning
39th International Conference on Machine Learning, ICML 2022, Baltimore, USA, 17 Jul 2022 - 23 Jul 2022 PMLR 10575-10585 (2022)

Abstract: In recent work, it has been shown that reinforcement learning (RL) is capable of solving a variety of problems at sometimes super-human performance levels. But despite continued advances in the field, applying RL to complex real-world control and optimisation problems has proven difficult. In this contribution, we demonstrate how to successfully apply RL to the optimisation of a highly complex real-world machine – specifically a linear particle accelerator – in an only partially observable setting and without requiring training on the real machine. Our method outperforms conventional optimisation algorithms in both the achieved result and time taken as well as already achieving close to human-level performance. We expect that such automation of machine optimisation will push the limits of operability, increase machine availability and lead to a paradigm shift in how such machines are operated, ultimately facilitating advances in a variety of fields, such as science and medicine among many others.

Contributing Institute(s):

Research Program(s):