Learning to Hold Plasma: Safe Reinforcement Learning for Tokamak Control
RL has shaped tokamak plasmas in simulation. Can it hold hard configurations more robustly than classical control without ever violating a…
article · by @science_geek
Thesis. Reinforcement-learning controllers can hold difficult plasma configurations with less control effort and more robustness to disturbances than classical control, but only a safety-constrained formulation keeps them inside the hard limits a device requires; a surrogate-trained, high-fidelity-validated pipeline can demonstrate both.