Measuring the electron temperature and identifying plasma detachment using machine learning and spectroscopy

• Published in: Review of scientific instruments 2021-04, Vol.92 (4), p.043520-043520
• Main Authors: Samuell, C. M., Mclean, A. G., Johnson, C. A., Glass, F., Jaervinen, A. E.
• Format: Article
• Language: English
• Subjects: Classifiers; Electron energy; Electrons; Emission spectra; Low cost; Low temperature; Machine learning; Neural networks; Scientific apparatus & instruments; Spectrometers; Spectrum analysis; Temperature; Thomson scattering; Tokamak devices; Ultraviolet emission
• ISSN: 0034-6748; 1089-7623
• DOI: 10.1063/5.0034552

A machine learning approach has been implemented to measure the electron temperature directly from the emission spectra of a tokamak plasma. This approach utilized a neural network (NN) trained on a dataset of 1865 time slices from operation of the DIII-D tokamak using extreme ultraviolet/vacuum ultraviolet emission spectroscopy matched with high-accuracy divertor Thomson scattering measurements of the electron temperature, Te. This NN is shown to be particularly good at predicting Te at low temperatures (Te < 10 eV) where the NN demonstrated a mean average error of less than 1 eV. Trained to detect plasma detachment in the tokamak divertor, a NN classifier was able to correctly identify detached states (Te < 5 eV) with a 99% accuracy (an F1 score of 0.96) at an acquisition rate 10× faster than the Thomson scattering measurement. The performance of the model is understood by examining a set of 4800 theoretical spectra generated using collisional radiative modeling that was also used to predict the performance of a low-cost spectrometer viewing nitrogen emission in the visible wavelengths. These results provide a proof-of-principle that low-cost spectrometers leveraged with machine learning can be used to boost the performance of more expensive diagnostics on fusion devices and be used independently as a fast and accurate Te measurement and detachment classifier.