Identification of Magnetic Field Errors in Synchrotrons based on Deep Lie Map Networks

TL;DR

This paper introduces deep Lie map networks, a data-driven machine learning approach to identify magnetic field errors in synchrotrons in parallel, improving efficiency and accuracy over traditional sequential methods.

Contribution

It presents a novel deep learning method that detects magnetic field errors in multiple accelerator sections simultaneously, reducing beam time and enhancing model precision.

Findings

Successfully inferred error locations and strengths in simulated data

Enabled parallel detection of gradient and sextupole errors

Improved accelerator model accuracy for better control and resonance compensation

Abstract

Magnetic field errors pose a limitation in the performance of synchrotrons, as they excite non-systematic resonances, reduce dynamic aperture and may result in beam loss. Their effect can be compensated assuming knowledge of their location and strength. Established identification procedures are based on orbit response matrices or resonance driving terms. While they sequentially build a field error model for subsequent accelerator sections, a method detecting field errors in parallel could save valuable beam time. We introduce deep Lie map networks, which enable construction of an accelerator model including multipole components for the magnetic field errors by linking charged particle dynamics with machine learning methodology in a data-driven approach. Based on simulated beam-position-monitor readings for the example case of SIS18 at GSI, we demonstrate inference of location and strengths of gradient and sextupole errors for all accelerator sections in parallel. The obtained refined accelerator model may support setup of corrector magnets in operation to allow more precise control over tunes, chromaticities and resonance compensation.