Scientists have developed a new machine learning platform that makes algorithms that control beams
Daniele Filippetto and his colleaguesof Lawrence Berkeley National Laboratory (Berkeley Lab) have developed a facility to automatically compensate for changes in real time in accelerator beams and other components such as magnets.
Research in this area will influenceacross many particle accelerator applications, ranging from autonomous operations in industrial and medical settings to precision enhancement in scientific applications such as linear colliders and ultrafast free electron lasers.
Filippetto's work now focuses on harnessing the power and prediction of machine learning tools to improve the overall stability of particle beams.
“If you can predict the properties of a beam withaccuracy exceeding their fluctuations, it is easy to use this prediction to improve the performance of the accelerator,” the scientist noted. “Knowledge of the key parameters of the beam in real time will have a huge impact on the final accuracy of the experiments.”
The method has already been demonstrated on an acceleratorfor High Repetition Rate Electron Scattering (HiRES) at Berkeley Lab in collaboration with researchers at Los Alamos National Laboratory and the University of California, Los Angeles. The main application of the HiRES beamline is to conduct structural dynamics experiments with new quantum materials.
The device has contributed to numerous scientificdiscoveries such as conducting the world's first ultrafast electron diffraction studies of the optical melting of tantalum ditelluride, a material with interesting and potentially useful properties. This new machine is now demonstrating its usefulness for developing new control methods for a wide class of accelerators.
Read more:
Physicists have found a universal "clock" in space: they are more accurate than atomic
Archaeologists have found drawings of creepy people with huge heads: who they were
The James Webb telescope took the first picture of Jupiter: it shows 9 moving targets at once