Bayesian optimization of a free-electron laser

TL;DR

This paper presents a Bayesian optimization method for tuning free-electron laser components, significantly improving efficiency over existing methods by using Gaussian processes and probabilistic modeling.

Contribution

The paper introduces a novel Bayesian optimization approach for free-electron laser tuning, incorporating Gaussian processes and beam transport models for enhanced efficiency.

Findings

Outperforms existing optimizers in sample efficiency

Uses Gaussian processes for probabilistic modeling of machine response

Incorporates beam transport correlations to improve tuning

Abstract

The Linac Coherent Light Source changes configurations multiple times per day, necessitating fast tuning strategies to reduce setup time for successive experiments. To this end, we employ a Bayesian approach to transport optics tuning to optimize groups of quadrupole magnets. We use a Gaussian process to provide a probabilistic model of the machine response with respect to control parameters from a modest number of samples. Subsequent samples are selected during optimization using a statistical test combining the model prediction and uncertainty. The model parameters are fit from archived scans, and correlations between devices are added from a simple beam transport model. The result is a sample-efficient optimization routine, which we show significantly outperforms existing optimizers.