Using Convolutional Neural Networks to detect Edge Localized Modes in DIII-D from Doppler Backscattering measurements

TL;DR

This paper develops a neural network-based method to detect edge localized modes in tokamak plasmas using Doppler backscattering data, achieving high accuracy and demonstrating broad applicability for plasma diagnostics.

Contribution

The study introduces a neural network approach for ELM detection from DBS data, validated against D-alpha measurements, with high performance on DIII-D shots.

Findings

Achieved a high F1-score of 0.93 on similar shots.

Demonstrated neural network effectiveness across different ELM types.

Showed feasibility of applying neural networks to DBS diagnostics.

Abstract

In H-mode tokamak plasmas, the plasma is sometimes ejected beyond the edge transport barrier. These events are known as edge localized modes (ELMs). ELMs cause a loss of energy and damage the vessel walls. Understanding the physics of ELMs and by extension, how to detect and mitigate them, is an important challenge. In this paper, we focus on two diagnostic methods $\unicode{x2013}$ D-alpha spectroscopy and Doppler backscattering (DBS). The former detects ELMs by measuring Balmer alpha emission while the latter uses microwave radiation to probe the plasma. DBS has the advantage of having higher temporal resolution and robustness to damage. These advantages of DBS diagnostics may be beneficial for future operational tokamaks and thus data processing techniques for DBS should be developed in preparation. In sight of this, we explore the training of neural networks to detect ELMs from DBS data, using D-alpha data as the ground truth. With shots found in the DIII-D database, the model is trained to classify each time step based on the occurrence of an ELM event. The results are promising. When tested on shots similar to those used for training, the model is capable of consistently achieving a high f1-score of 0.93. This score is a performance metric for imbalanced datasets that ranges between 0 and 1. We evaluate the performance of our neural network on a variety of ELMs $\unicode{x2013}$ grasssy, suppressed, and wide pedestal $\unicode{x2013}$ finding broad applicability. Beyond ELMs, our work demonstrates the wider feasibility of applying neural networks to data from DBS diagnostics.