ORCID

0000-0002-3475-2871 (Schram), 0000-0003-4162-0276 (Colen)

Document Type

Article

Publication Date

2025

DOI

10.1088/2632-2153/adc221

Publication Title

Machine Learning: Science and Technology

Volume

6

Issue

2

Pages

025018 (1-16)

Abstract

Particle accelerator operation requires simultaneous optimization of multiple objectives. Multi-objective optimization (MOO) is particularly challenging due to trade-offs between the objectives. Evolutionary algorithms, such as genetic algorithms (GAs), have been leveraged for many optimization problems, however, they do not apply to complex control problems by design. This paper demonstrates the power of differentiability for solving MOO problems in particle accelerators using a deep differentiable reinforcement learning (DDRL) algorithm. We compare the DDRL algorithm with model-free reinforcement learning (MFRL), GA, and Bayesian optimization (BO) for simultaneous optimization of heat load and trip rates in the continuous electron beam accelerator facility. The underlying problem enforces strict constraints on both individual states and actions as well as cumulative (global) constraints on energy requirements of the beam. Using historical accelerator data, we develop a physics-based surrogate model which is differentiable and allows for back-propagation of gradients. The results are evaluated in the form of a Pareto-front with two objectives. We show that the DDRL outperforms MFRL, BO, and GA on high dimensional problems.

Rights

© 2025 The Authors.

Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 International (CC BY 4.0) License. Any further distribution of this work must maintain attribution to the authors and the title of the work, journal citation and DOI.

Data Availability

Article states: "The data cannot be made publicly available upon publication due to legal restrictions preventing unrestricted public distribution. The data that support the findings of this study are available upon reasonable request from the authors."

Original Publication Citation

Rajput, K., Schram, M., Edelen, A., Colen, J., Kasparian, A., Roussel, R., Carpenter, A., Zhang, H., & Benesch, J. (2025). Harnessing the power of gradient-based simulations for multi-objective optimization in particle accelerators. Machine Learning: Science and Technology, 6(2), 1-16, Article 025018. https://doi.org/10.1088/2632-2153/adc221

Download

Share

 
COinS