Robust and Model Predictive Control for Boundary Control Systems

In this thesis, robust and model predictive control are considered for boundary control systems. In terms of robust control, the existing results, especially the internal model principle, are generalized to cover this class of systems. The concept of approximate robust regulation for boundary control systems is presented, as, due to the internal model principle, in practice it is not possible to construct an exact robust regulating controller if the output space of the controlled system is infinite-dimensional. A practical controller design is presented to achieve robust regulation in this approximate sense.

Model predictive control (MPC) is considered for the class of regular linear systems which includes regular boundary control systems. The continuous-time system is approximated by a discrete-time one by using the Cayley-Tustin transform, and MPC is considered for the discrete-time system. Stability and optimality are proved for the proposed discrete-time MPC designs, which extends the corresponding finitedimensional MPC designs to the class of regular linear systems.