Dortmund 2021 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

AKBP: Arbeitskreis Beschleunigerphysik

AKBP 6: Diagnostics, Control and Instrumentation I

AKBP 6.8: Vortrag

Mittwoch, 17. März 2021, 15:45–16:00, AKBPa

Sensitivity Analysis of Beam-influencing Parameters at the S-DALINAC Using Surrogate Models* — •Dominic Schneider, Michaela Arnold, Jonny Birkhan, Norbert Pietralla, and Felix Schließmann — Institut für Kernphysik, TU Darmstadt, Darmstadt

Machine learning methods provide a significant potential for the optimized operation of complex machinery, such as particle accelerators. In this contribution, the first application of so-called surrogate models to the electron accelerator S-DALINAC [1] will be discussed. This machine learning technique, based on polynomial fitting, gives not only access to predict future behavior based on training data, but also an extensive set of characteristics that can be extracted by analyzing the trained model. The talk will focus on a series of measurements that have been performed at the S-DALINAC in order to investigate the behavior and correlations of beam-influencing elements on the one hand and the performance of surrogate models on the other hand. Particularly, the global sensitivity analysis as well as Sobol indices will be discussed.

[1] N. Pietralla, Nuclear Physics News, Vol. 28, No. 2, 4 (2018)

*Work supported by DFG through GRK 2128