To content

Machine Learning (ML) Applications

Accelerator facilities are equipped with highly complex control systems with several thousand parameters, some of which must be regulated with high accuracy and within small time scales. Today, artificial intelligence (AI) methods, in particular machine learning (ML) algorithms, support the automatic control, monitoring and diagnosis of accelerator operation. Due to its high availability for accelerator physics studies, the electron storage ring facility DELTA offers an excellent test environment for the development of novel innovative AI-based control and optimization methods. The applications cover a wide range of tasks.

Typical workflow (steps 1 to 6) in the realization of ML-based applications at the electron storage ring facility DELTA.

The following ML projects have been implemented so far:

  • Self-regulating trajectory correction of the stored electron beam.
  • Feedback systems to control the working points and chromaticity values of the storage ring.
  • Optimization of the electron transfer rate from the pre-accelerator to the storage ring (injection optimization).
  • Spectral analysis of CHG (Coherent Harmonic Generation) radiation.

More detailed information can be found under the menu items below.