Poloidal Flux Profile Reconstruction From Pointwise Measurements via Extended Kalman Filtering in the DIII-D Tokamak

H. Wang, J.E. Barton and E. Schuster

IEEE Multi-conference on Systems and Control

Sydney, Australia, September 21-23, 2015

Abstract

The accuracy of the internal states of a tokamak, which usually cannot be measured directly, is of crucial importance for feedback control of the plasma dynamics. A first-principles-driven plasma response model could provide an estimation of the internal states given the boundary conditions on the magnetic axis and at plasma boundary. However, the estimation would highly depend on initial conditions, which may not always be known, disturbances, and non-modeled dynamics. In this work, a closed-loop state observer for the poloidal magnetic flux is proposed based on a very limited set of real-time measurements by following an Extended Kalman Filtering (EKF) approach. Comparisons between estimated and measured magnetic flux profiles are carried out for several discharges in the DIII-D tokamak. The experimental results illustrate the capability of the proposed observer in dealing with incorrect initial conditions and measurement noise.