Robust model predictive control of a nuclear power plant pressurizer subsystem

Robust model predictive control of an industrial pressurizer is presented in this paper. The physical model of the pressurizer is based on first engineering principles and the model parameters have been previously identified from measured data. To satisfy the hard constraints on the state variables and the input even in the presence of disturbances, the so-called single policy robust model predictive control method is applied. The maximal admissible level set, the disturbance invariant set and the terminal sets are determined for the system. Simulation results show that the proposed controller satisfies all the requirements and shows good time-domain behavior.