Plasma state identification in the COMPASS tokamak

0571890 - ÚFP DATA Scientific data      2023
Zorek, M. - Škvára, V. - Šmídl, V. - Pevný, T. - Seidl, Jakub - Grover, Ondřej

Plasma state identification in the COMPASS tokamak.

Description: The dataset contains selected diagnostic measurements from the COMPASS tokamak and manually created labels of plasma state in the form of csv files. The dataset defines 4 plasma states: L-mode, H-mode, ELM leading edge and ELM trailing edge. The data is intended to be used for training and testing deep neural networks in the task of plasma state identification. The dataset is divided into two parts, supervised (labeled) and unsupervised (unlabeled), each containing 20 COMPASS discharges. Within the unsupervised dataset there are 11 labeled discharges that can be used as a test set.

Keywords : COMPASS * tokamak * h-mode * l-mode
OECD category: Fluids and plasma physics (including surface physics)
DOI: https://doi.org/10.57680/asep.0571890
Handle: https://hdl.handle.net/11104/0342782
Depositor: admin
Date of release: 18.5.2023
 
ASEP publication:
Semi-supervised deep networks for plasma state identification