Optimizing a Superconducting Radiofrequency Gun Using Deep Reinforcement Learning

TL;DR

This paper presents a novel reinforcement learning method to optimize superconducting radiofrequency gun parameters, using neural network approximations to reduce computation time and outperform traditional optimization techniques.

Contribution

It introduces a reinforcement learning approach combined with neural network approximations for efficient optimization of complex superconducting gun parameters.

Findings

Reinforcement learning outperforms traditional methods in optimization efficiency.

Neural network approximation significantly reduces simulation computation time.

The approach achieves high accuracy in beam property optimization.

Abstract

Superconducting photoelectron injectors are a promising technique for generating high brilliant pulsed electron beams with high repetition rates and low emittances. Experiments such as ultra-fast electron diffraction, experiments at the Terahertz scale, and energy recovery linac applications require such properties. However, optimization of the beam properties is challenging due to the high amount of possible machine parameter combinations. In this article, we show the successful automated optimization of beam properties utilizing an already existing simulation model. To reduce the amount of required computation time, we replace the costly simulation by a faster approximation with a neural network. For optimization, we propose a reinforcement learning approach leveraging the simple computation of the derivative of the approximation. We prove that our approach outperforms common optimization methods for the required function evaluations given a defined minimum accuracy.