Tokamak rotation profile optimization through Model Predictive Control

Optimizing the plasma rotation profile, via neutral beam injection (NBI) and 3D magnetic field-induced neoclassical toroidal viscosity (NTV) torque, can be vital for ensuring robust pedestal performance and tearing stability in various operating scenarios. Metrics for optimizing the rotation profile include maximizing the rotation shear at various rational surfaces along the safety factor profile. A model predictive control (MPC) algorithm which controls NBI and NTV torque during TRANSP simulations, in which the rotation profile is predicted based on a fixed momentum diffusivity, is used to investigate how neutral beams and 3D coils may be actuated to tailor the plasma rotation profile as desired. Simplified models for NBI and NTV torque are provided in situ during the TRANSP simulation by the NUBEAM and Generalized Perturbed Equilibrium Code (GPEC) numerical codes, respectively. The MPC algorithm involves optimizing a simplified version of the toroidal momentum equation over a predetermined control horizon. Other objectives for optimizing the rotation profile are also discussed.