Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
P: Fachverband Plasmaphysik
P 15: Helmholtz Graduate School II
P 15.2: Poster
Dienstag, 14. März 2017, 16:30–18:30, HS Foyer
Neural Net Applications for Plasma Edge Analysis in Wendelstein 7-X — •Marko Blatzheim1,2, Daniel Böckenhoff1, Hauke Hölbe1, Thomas Sunn Pedersen1, and Roger Labahn2 — 1Max-Planck-Institut für Plasmaphysik, Greifswald, Deutschland — 2Universität Rostock, Rostock, Deutschland
Neural nets are powerful tools and due to recent improvements in computer performance and more complex mathematical approaches they are the state of the art in various applications, e.g. pattern recognition, machine translation or human-level control. Wendelstein 7-X (W7-X) is a fully optimized stellarator with the main goal to demonstrate steady state capability of fusion reactors. The plasma edge targets so-called divertors which ensure to avoid damage at other first-wall components and to reduce plasma impurities. Therefore, they themselves are exposed to a high heat load which can be observed by infrared cameras. We are using neural nets with the purpose to analyze images of the plasma-divertor-interaction in real time. Convolutional neural nets are trained using simulated plasma data. They are a promising approach to predict different plasma properties. In the future these neural nets will be trained to evaluate possible critical states and find parameter configurations to avoid them.