Analysis and LQ-Optimal Control of Distributed Parameter Systems

This thesis deals with the synthesis of optimal control laws with a view to regulate the temperature and the reactant concentration of a nonisothermal plug flow reactor model. Several tools of linear and semilinear infinite-dimensional system theory are extended and/or developed, and applied to this model. On the one hand, the concept of asymptotic stability is studied for a class of infinite-dimensional semilinear Banach state- space systems. Asymptotic stability criteria are established, which are based on the concept of strictly m-dissipative operator. This theory is applied to a nonisothermal plug flow reactor. On the other hand, the concept of optimal Linear-Quadratic (LQ) feedback is studied for class of infinite-dimensional linear systems. This theory is applied to a linearized plug flow reactor model in order to design an LQ optimal feedback controller. Then the resulting nonlinear closed-loop system performances are analyzed. Finally this control design strategy is extended to a large class of first-order hyperbolic PDE¿s systems.