Speaker
Description
Pulse pile-up limits the count rate, timing precision, and energy resolution of radiation detection systems in accelerator beam instrumentation and downstream experiments. We present the project status of a machine-learning based pile-up recovery system designed for real-time particle counting and time of arrival determination at pulse rates exceeding $10^7$~particles/s. The convolutional neural network (CNN) architecture utilized in this project is trained on labelled scintillator data to identify pulses in the piled-up waveforms. This development is primarily aimed at beam spill characterization at GSI/FAIR using plastic scintillators, however the concept presented is general-purpose and could be applicable to any radiation detector. First model implementation is performed on the FPGA onboard a commercial digitizer and its performance is compared against the single threshold leading-edge discriminator.
| In which format do you inted to submit your paper? | LaTeX |
|---|---|
| Preprint marking on your proceeding paper | I wish my paper to be marked as preprint. |