Skip to content

Demos of Cheetah being used for various applications presented in "Cheetah: Bridging the Gap Between Machine Learning and Particle Accelerator Physics with High-Speed, Differentiable Simulations"

License

Notifications You must be signed in to change notification settings

desy-ml/cheetah-demos

Repository files navigation

codestyle License: GPL v3

Cheetah Demos

This repository contains a collection of demos accompanying the Cheetah high-speed, differentiable beam dynamics simulation Python package.

For more information, see the paper where these demos were first introduced: Bridging the Gap Between Machine Learning and Particle Accelerator Physics with High-Speed, Differentiable Simulations.

Finding your way around

  • benchmark: Various speed benchmarks for Cheetah and other simulation tools.
  • bo_prior: Example of using a differentiable Cheetah model as a prior for Bayesian optimisation on a particle accelerator to improve tuning performance.
  • neural_network_space_charge_quad: Implementation of a modular neural network surrogate model for high-speed computation of space charge effects through a quadrupole magnet.
  • reinforcement_learning: Data and plotting code for example tuning performed by a neural network policy trained with reinforcement learning using a Cheetah simulation environment. The full RL example can be found in Learning-based Optimisation of Particle Accelerators Under Partial Observability Without Real-World Training.
  • system_identification: Example of using Cheetah with gradient-based optimisation to identify the parameters of a particle accelerator model from noisy measurements.
  • tuning: Example of using Cheetah with gradient-based optimisation to tune a particle accelerator subsection to a desired working point.

Cite this repository

Please cite the original paper that these demos were introduced in:

@article{kaiser2024cheetah,
    title        = {Bridging the gap between machine learning and particle accelerator physics with high-speed, differentiable simulations},
    author       = {Kaiser, Jan and Xu, Chenran and Eichler, Annika and Santamaria Garcia, Andrea},
    year         = 2024,
    month        = {May},
    journal      = {Phys. Rev. Accel. Beams},
    publisher    = {American Physical Society},
    volume       = 27,
    pages        = {054601},
    doi          = {10.1103/PhysRevAccelBeams.27.054601},
    url          = {https://link.aps.org/doi/10.1103/PhysRevAccelBeams.27.054601},
    issue        = 5,
    numpages     = 17
}

About

Demos of Cheetah being used for various applications presented in "Cheetah: Bridging the Gap Between Machine Learning and Particle Accelerator Physics with High-Speed, Differentiable Simulations"

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •  

Languages