Suppressing Plasma Instability Through Constrained Optimization
Li Wang
University of Minnesota, Twin Cities
Maintaining the stability and shape of a plasma is a crucial task in fusion energy. This is often challenging as plasma systems tend to be naturally unstable and kinetic e?ects can play an important role in the behavior of the instabilities. In this talk, I will introduce PDE-constrained optimization formulation that uses a kinetic description of plasma dynamics, particularly the Vlasov–Poisson system as the governing constraint. I will then discuss strategies to reduce computational costs through moment control. To further address the challenge of real-time control of plasma instabilities over long time horizons, we develop a dynamic feedback control strategy. This involves constructing an operator that maps state perturbations to an external control field. This operator is either approximated by a simple neural network, or directly constructed via energy estimates and a cancellation-based control approach that neutralizes the destabilizing components of the electric field.