Speaker
Mark Daniel Boyer
Description
See the full Abstract at http://ocs.ciemat.es/EPS2018ABS/pdf/P4.1084.pdf
Real-time capable neural network approximation of NUBEAM for use in
the NSTX-U control system
M.D. Boyer1, S. Kaye1, D. Liu2, K. Erickson1, W. Heidbrink2,
O. Menegheni3, S.A. Sabbagh4
1 Princeton Plasma Physics Laboratory, Princeton, NJ, USA
2University of California-Irvine, Irvine, CA, USA
3General Atomics, San Diego, CA, USA
4Columbia University, New York, NY, USA
Present-day and next step tokamaks will require precise control of plasma conditions,
including the spatial distribution of rotation and current profiles, in order to optimize
performance and avoid physics and operational constraints. The coupled nonlinear
dynamics of equilibrium profiles and the complex effects of actuators on the equilibrium
evolution motivates embedding physics-based models within real-time control algorithm
designs. Due to the important role of beam heating, current drive, and torque in
establishing scenario performance and stability, a high-fidelity beam model suitable for
use in real-time applications is desired. This work describes a neural network that has
been developed to enable rapid evaluation of the beam heating, torque, and current drive
profiles based on measured equilibrium profiles. The training and testing database has
been generated from the NUBEAM calculations output from interpretive TRANSP
analysis of shots from the 2016 NSTX-U campaign, including scans of Zeff and fast ion
diffusivity. Neural network predictions made for the testing data demonstrate the ability
of the model to generalize and accurately reproduce NUBEAM calculated profiles and
scalar quantities. Results of hardware-in-the-loop simulations of the model within the
NSTX-U plasma control system will be presented, along with plans and progress toward
application of the neural network for accelerated offline analysis and real-time control.
* Work supported by U.S.D.O.E. Contract No. DE-AC02-09CH11466.
