Tokamak plasmas experience a large number of macroscopic and microscopic instabilities, which produce losses of particles and energy or, in the worst cases, the disruption of the plasma altogether. These have to be stabilized through active feedback control, which implies the use of appropriate diagnostics to measure the physical event, processing algorithms, to separate the background noise and synthesize appropriate control actions, and actuators, to feed back to the plasma. My work consists of thinking and implementing new algorithms and hardware for control purposes. Recently, my research has concentrated on developing a simulation code for axisymmetric control analysis, where the model of the plasma is intentionally simplified, but the vertical stability of elongated plasmas is correctly reproduced. Boundary conditions, such as the saturation of the poloidal field power supplies, are also accounted in the model. The ultimate goal is to provide the control engineer with a tool for testing new ideas (adaptive control for off-normal events, control in the presence of actuator saturation, etc.) and the physics operator with a fast software to test the plasma stability before real discharges. The simulator results have been compared with real discharges and have shown good agreement, when the setup follows the nominal configuration of Alcator C-Mod. In particular, the marginal stability of fairly elongated plasmas has been correctly reproduced, and a fine investigation of PID gains is ongoing to improve the control of those plasmas. In general, research efforts for improving the reliability and effectiveness of tokamak control are essential for the next generation thermonuclear reactors, including the international experiment ITER, where minor control failures could result in poor performance or even machine damage.

http://web.mit.edu/mferrara/www/home/home.html

77 Massachusetts Avenue, NW16, Cambridge, MA 02139, info@psfc.mit.edu