Real time equilibrium reconstruction for control of the discharge in the DIII-D tokamak

Optimum performance of a tokamak discharge requires accurate feedback control of many of the discharge parameters. For this to be possible, the values of these parameters must be accurately measured. The values of many discharge parameters, such as shape and safety factor profile, are not directly measured but can be evaluated from the available diagnostic data: magnetic field and flux measurements, for example. The most complete evaluation comes from a least squares fit of the diagnostic data to the Grad-Shafranov model that describes the force balance of the tokamak equilibrium, while allowing for a distributed current source. This full reconstruction of the equilibrium has normally been performed offline using a computation-intensive fitting code such as EFIT. This paper provides an introduction to a practical method for performing an equilibrium reconstruction in real time for arbitrary time-varying discharge shapes and current profiles. A detailed description of the algorithm is given in Ref. 2. An approximate solution to the Grad-Shafranov equilibrium relation is found which best fits the diagnostic measurements so that an equilibrium solution consistent with force balance, expressed in terms of the spatial distributions of the toroidal current density and poloidal flux, is available in real time for accurate evaluation of the discharge parameters. The algorithm is very close to that of EFIT and is executed on a time scale fast enough for control of the DIII-D tokamak.