Doctor of Philosophy
Tokamaks are devices whose final purpose is obtaining energy by means of nuclear fusion reactions. To achieve such purpose, a gas is injected into the tokamak's torus-shaped chamber and heated up to extremely high temperatures, giving birth to a plasma. When the necessary conditions of temperature, density, and confinement time are achieved, virtually inexhaustible energy can be produced in a tokamak.
The main contribution of this dissertation is the development of integrated control strategies for tokamak plasmas. The development of integrated control architectures is necessary for tokamaks to become efficient and commercially competitive power plants. Because a tokamak plasma is a highly nonlinear, coupled dynamical system, the great diversity of complex control problems that coexist in a tokamak are indeed closely interrelated. However, this variety of control problems must be tackled by means of a limited number of actuators. A functional design for integrated tokamak-control architectures should employ multi-input multi-output controllers to simultaneously regulate as many plasma variables as possible with the available actuators. Supervisory and exception handling systems that monitor the plasma state arise as a necessity to ensure a safe tokamak-operation. Finally, actuator sharing and management capabilities should also exist in order to utilize the available actuators in an optimal way. Various control problems are tackled in this dissertation, including kinetic, magnetic, and instability control problems. Control-oriented, physics-based models that characterize some specific aspects of the plasma dynamics have been employed to develop new control-oriented simulation codes and integrated-control solutions that employ nonlinear, robust control techniques and optimization-based actuator-management strategies. Some of those control solutions have been experimentally tested in the DIII-D tokamak.
Pajares, Andres, "Integrated Control in Tokamaks using Nonlinear Robust Techniques and Actuator Sharing Strategies" (2019). Theses and Dissertations. 5728.