Towards a Robust Adaptive Digital Twin for Fusion Applications

November 17, 2025·
admin
· 1 min read
Image credit: Kishan
Abstract
Developing accurate and robust digital twins for fusion plasma control is essential for advancing fusion energy research. We present methods for creating adaptive digital twins that can handle non-stationary data and provide uncertainty estimates. This work demonstrates how advanced machine learning techniques can improve model predictions in the face of changing accelerator conditions and plasma dynamics.
Date
November 17, 2025 2:00 PM — 2:15 PM
Location

San Diego, CA

San Diego, CA,

events

Presentation at the 67th Annual Meeting of the APS Division of Plasma Physics in San Diego, 2025.

Talk Machine Learning Digital Twins Fusion Science Uncertainty Quantification Adaptive Learning
Authors
admin