Machine-learning in Secondary Emission Monitor (SEM) optimisation
The Super Proton Synchrotron (SPS) at CERN, is the last step in the acceleration of particles at CERN before they enter the Large Hadron Collider (LHC), with energies of 450 GeV. As well providing beams for the LHC, the SPS also provides beams for other CERN experiments, such as AWAKE, NA61/SHINE, NA62 and COMPASS.
Some of these experiments require slow extraction. At the SPS slow extraction lines Secondary Emission Monitors (SEMs) are used for an in detail characterisation of the primary beam. They work by detecting secondary electrons, emitted as the primary beam passes through the mesh or foil of the SEM.
The research conducted in this project is focused on improving the absolute calibration and calibration stability of the SEM foils used in the slow extraction lines of the SPS. The aim is to create new solutions for online calibration, comparing SEM data with that of Optical Transition Radiation (OTR) and Cherenkov detectors, to optimise the electrodes/grids of the SEMs to improve secondary emission, and to employ machine learning techniques to boost the data analysis.