Full Shot Predictions for the DIII-D Tokamak via Deep Recurrent Networks
This work addresses the challenge of plasma control and understanding in nuclear fusion for energy applications, but it is incremental as it applies an existing deep learning method to new tokamak data.
The researchers tackled the problem of predicting the full time evolution of plasma discharges in the DIII-D tokamak using a data-driven approach, resulting in a deep recurrent network model that can forecast entire shots, with an investigation into how training and inference procedures impact prediction quality and calibration.
Although tokamaks are one of the most promising devices for realizing nuclear fusion as an energy source, there are still key obstacles when it comes to understanding the dynamics of the plasma and controlling it. As such, it is crucial that high quality models are developed to assist in overcoming these obstacles. In this work, we take an entirely data driven approach to learn such a model. In particular, we use historical data from the DIII-D tokamak to train a deep recurrent network that is able to predict the full time evolution of plasma discharges (or "shots"). Following this, we investigate how different training and inference procedures affect the quality and calibration of the shot predictions.