Enhancing disruption prediction through Bayesian neural network in KSTAR

• Published in: Plasma physics and controlled fusion Vol. 66; no. 7; pp. 75001 - 75021
• Main Authors: Kim, Jinsu, Lee, Jeongwon, Seo, Jaemin, Ghim, Young-Chul, Lee, Yeongsun, Na, Yong-Su
• Format: Journal Article
• Language: English
• Published: IOP Publishing 01.07.2024
• Subjects: Bayesian neural network; deep learning; disruption; tokamak; uncertainty estimation
• ISSN: 0741-3335; 1361-6587
• DOI: 10.1088/1361-6587/ad48b7

In this research, we develop a data-driven disruption predictor based on Bayesian deep probabilistic learning, capable of predicting disruptions and modeling uncertainty in KSTAR. Unlike conventional neural networks within a frequentist approach, Bayesian neural networks can quantify the uncertainty associated with their predictions, thereby enhancing the precision of disruption prediction by mitigating false alarm rates through uncertainty thresholding. Leveraging 0D plasma parameters from EFIT and diagnostic data, a temporal convolutional network adept at handling multi-time scale data was utilized. The proposed framework demonstrates proficiency in predicting disruptions, substantiating its effectiveness through successful applications to KSTAR experimental data.