Göttingen 2025 – wissenschaftliches Programm

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

P: Fachverband Plasmaphysik

P 17: Poster Session II

P 17.49: Poster

Mittwoch, 2. April 2025, 16:15–18:15, ZHG Foyer 1. OG

Fast Surrogate Modeling of Radio-Frequency Minority Heating at ASDEX Upgrade — •Michael Sieben, Markus Weiland, Roberto Bilato, and ASDEX - Team — Max-Planck-Institute for Plasma Physics

Auxiliary heating is essential to achieve the burning plasma conditions in fusion reactors. Radio-frequency waves in the Ion Cyclotron Range of Frequency (ICRF) are one of the possible auxiliary heating system in present devices and planned for ITER. The combined full-wave and Fokker-Planck code package TORIC-SSFPQL [1] provides accurate heating profiles for transport modeling in view of discharge analysis and design. However, long computation times pose challenges for fast transport simulations and real-time applications. We present a fast neural network (NN) surrogate model that reproduces minority heating profiles in deuterium-hydrogen plasmas at the ASDEX Upgrade (AUG) research reactor with real-time capability and high accuracy - comparable to an approach reported for NSTX and WEST [2]. We highlight possible directions for future work that could improve model accuracy - particularly by accounting for second harmonic heating and integrating equilibrium parameters into the model, which has been found to play a relevant role at AUG.

[1] M. Brambilla and R. Bilato. Nuclear Fusion, 49(8):085004, 2009. [2] Á. Sánchez-Villar, Z. Bai, et al. Nuclear Fusion, 64(9):096039, 2024.

Keywords: ICRF; plasma heating; machine learning; neural network; surrogate modeling