Integrated magnetic and kinetic control of advanced tokamak plasmas on DIII-D based on data-driven models

Abstract

The first real-time profile control experiments integrating magnetic and kinetic variables were performed on DIII-D in view of regulating and extrapolating advanced tokamak scenarios to steady-state devices and burning plasma experiments. Device-specific, control-oriented models were obtained from experimental data using a generic two-time-scale method that was validated on JET, JT-60U and DIII-D under the framework of the International Tokamak Physics Activity for Integrated Operation Scenarios (Moreau et al 2011 Nucl. Fusion 51 063009). On DIII-D, these data-driven models were used to synthesize integrated magnetic and kinetic profile controllers. The neutral beam injection (NBI), electron cyclotron current drive (ECCD) systems and ohmic coil provided the heating and current drive (H&CD) sources. The first control actuator was the plasma surface loop voltage (i.e. the ohmic coil), and the available beamlines and gyrotrons were grouped to form five additional H&CD actuators: co-current on-axis NBI, co-current off-axis NBI, counter-current NBI, balanced NBI and total ECCD power from all gyrotrons (with off-axis current deposition). Successful closed-loop experiments showing the control of (a) the poloidal flux profile, Ψ(x), (b) the poloidal flux profile together with the normalized pressure parameter, βN, and (c) the inverse of the safety factor profile, , are described.

Control-oriented modeling of the plasma particle density in tokamaks and application to real-time density profile reconstruction

“…ASTRA [30], CRONOS [43]) are not directly suitable for the task of real-time density reconstruction, since their execution time generally exceeds the discharge duration. It has been shown in [2][3][4][5]39,40] that low-complexity 1D models can be used for reconstruction and control of the temperature and safety factor profiles.…”

“…Plasma control has expanded in recent years from control of bulk plasma quantities (such as total plasma current, average particle density and average temperature) to control of the spatial distributions of these quantities, e.g. the profiles of temperature, safety factor and rotation [2][3][4][5][6].…”

Novel aspects of plasma control in ITER

“…Substantial experimental and theoretical research has been done to address the physics understanding and models needed, and then to explore candidate approaches to algorithm design for the profile-tracking mission. This research has employed a wide variety of methods, including black or gray box system identification, [27][28][29] largely physics-based but simplified transport equations, 23,24,30,31 or adaptive techniques. 32 Because of the complexity of the physics involved in current profile dynamics, and the difficulty in predicting plasma conditions, much of this research has made use of black or gray box approaches to modeling.…”

DIII-D accomplishments and plans in support of fusion next steps

“…This starts with reliable passively stable discharges, as discussed in section II.A. In a second layer, control is being developed to help maintain the plasma in a passively stable state through techniques such as current profile control [16] and error field correction. However, if an instability does occur, DIII-D is developing various tools to actively control these.…”