Bayesian Optimization of a Laser-Plasma Accelerator

Jalas, Sören; Kirchen, Manuel; Messner, Philipp; Winkler, Paul; Hübner, Lars; Dirkwinkel, Julian; Schnepp, M.; Lehe, Rémi; Maier, Andreas R.

doi:10.1103/physrevlett.126.104801

Cited by 92 publications

(67 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Another very recent example of BO application is in Laser-Plasma research for tuning the "unknown parameters" of the new ionisation injection acceleration scheme both in numerical simulation and in the corresponding experiment. Optimal conditions permitted generation of stable electron beam with subpercent level of beam energy spread [139].…”

Section: Advanced Optimization Techniquesmentioning

confidence: 99%

Positron sources: from conventional to advanced accelerator concepts-based colliders

Chaikovska,

Chehab,

Kubytskyi

et al. 2022

Preprint

View full text Add to dashboard Cite

A: Positron sources are the key elements for the future and current lepton collider projects such as ILC, CLIC, SuperKEKB, FCC-ee, Muon Collider/LEMMA, etc., introducing challenging critical requirements for high intensity and low emittance beams in order to achieve high luminosity. In fact, due to their large production emittance and constraints given by the target thermal load, the main collider parameters such as the peak and average current, the emittances, the damping time, the repetition frequency and consequently the luminosity are determined by the positron beam characteristics. In this paper, the conventional positron sources and their main properties are explored for giving an indication to the challenges that apply during the design of the advanced accelerator concepts. The photon-driven positron sources as the novel approach proposed, primarily for the future linear colliders, are described highlighting their variety and problematic.

show abstract

Section: Advanced Optimization Techniquesmentioning

confidence: 99%

Positron sources: from conventional to advanced accelerator concepts-based colliders

Chaikovska,

Chehab,

Kubytskyi

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…With increasing complexity of the target and the acceleration dynamics it becomes ever more difficult to identify an optimum working point for the plasma accelerator and then to reach this working point reliably on a day-to-day basis. To address this problem, machine learning techniques such as Bayesian optimization have been applied to the LUX accelerator [ 61 ] to, first, identify an optimum accelerator setting using particle-in-cell (PIC) simulations. A key factor for this study was the recent advances in the PIC domain, that enabled an accurate modeling of the experiment conditions in combination with a fast execution time and numerical accuracy [ 62 , 63 ] .…”

Section: Desy Luxmentioning

confidence: 99%

Free electron lasers driven by plasma accelerators: status and near-term prospects

Emma

Tilborg

Aßmann

et al. 2021

High Pow Laser Sci Eng

View full text Add to dashboard Cite

Owing to their ultra-high accelerating gradients, combined with injection inside micrometer-scale accelerating wakefield buckets, plasma-based accelerators hold great potential to drive a new generation of free-electron lasers (FELs). Indeed, the first demonstration of plasma-driven FEL gain was reported recently, representing a major milestone for the field. Several groups around the world are pursuing these novel light sources, with methodology varying in the use of wakefield driver (laser-driven or beam-driven), plasma structure, phase-space manipulation, beamline design, and undulator technology, among others. This paper presents our best attempt to provide a comprehensive overview of the global community efforts towards plasma-based FEL research and development.

show abstract

“…In contrast to our contribution, they demonstrate a general control theoretical motivated feedback algorithm, capable of performing online multi-objective optimisation. Bayesian optimisation in combination with particle-in-cell simulations [11] is used to tune a plasma accelerator autonomously to the beam energy spread to the sub-percent at a given energy and intensity.…”

Section: Introductionmentioning

confidence: 99%

Fast, Efficient and Flexible Particle Accelerator Optimisation Using Densely Connected and Invertible Neural Networks

2021

View full text Add to dashboard Cite

Particle accelerators are enabling tools for scientific exploration and discovery in various disciplines. However, finding optimised operation points for these complex machines is a challenging task due to the large number of parameters involved and the underlying non-linear dynamics. Here, we introduce two families of data-driven surrogate models, based on deep and invertible neural networks, that can replace the expensive physics computer models. These models are employed in multi-objective optimisations to find Pareto optimal operation points for two fundamentally different types of particle accelerators. Our approach reduces the time-to-solution for a multi-objective accelerator optimisation up to a factor of 640 and the computational cost up to 98%. The framework established here should pave the way for future online and real-time multi-objective optimisation of particle accelerators.

show abstract

Bayesian Optimization of a Laser-Plasma Accelerator

Cited by 92 publications

References 29 publications

Positron sources: from conventional to advanced accelerator concepts-based colliders

Positron sources: from conventional to advanced accelerator concepts-based colliders

Free electron lasers driven by plasma accelerators: status and near-term prospects

Fast, Efficient and Flexible Particle Accelerator Optimisation Using Densely Connected and Invertible Neural Networks

Contact Info

Product

Resources

About