Improved reconstruction of plasma profiles and equilibrium through dynamic state estimation: from TCV and AUG to ITER

Combining dynamic models and measurements into a consistent plasma state estimate is an important challenge for FPP/DEMO reactors, with limited diagnostic coverage. In this work, the RAPTOR fast transport solver is used as a dynamic state observer, applying an Extended Kalman Filter (EKF) to assimilate measurements into the simulation. Using the estimated current profile as constraint for equilibrium reconstruction, consistent transport and equilibrium solutions are achieved. Uncertain quantities, such as transport coefficients and Z_eff, previously assigned on an ad-hoc basis, are now consistently estimated. The EKF can be employed for both post-discharge analysis and for real-time plasma control and monitoring.

This contribution presents the real-time implementation of the RAPTOR EKF in the plasma control system of TCV and ASDEX Upgrade, which enables estimates of electron temperature and density and plasma current density, based on the available real-time measurements. The current density profile reconstruction is improved, including sawtooth dynamics, as verified through independent measurements, opening a pathway towards more accurate MHD mode control.

The dynamic state observer is tested for an ITER discharge with synthetic measurements generated from a DINA-JINTRAC simulation and improves q profile reconstructions in the absence of direct measurements. The RAPTOR transport EKF is coupled with the CREATE-NL free boundary equilibrium EKF, including vessel currents estimates.