An application of selective modal analysis to tokamak modeling and control

In this paper, the problem of model order reduction for tokamak devices is considered. Due to the complexity of the machine, the available models of the plasma behavior and its electromagnetic coupling with the active and metallic structures have to be very detailed to be usefully employed for simulation and control design. Such a level of detail consequently yields linearized models of the system of very high order. Having to deal with such high-order models makes the use of many control schemes quite complicated. To cope with this problem, the natural strategy is to derive models of reduced order, which reproduce with sufficient accuracy the full-order model behavior. The use of a reduction scheme which is based on selective modal analysis (SMA) is proposed. Its basic property is that the physical meaning of the state variables is preserved in the reduced model. This fact can be usefully employed both in the analysis of the system properties and for control design. In particular, the design of an LQG controller for the ITER machine is discussed.