Input-output linearizing nonlinear model predictive control (Q4374764)

The continuous-time feedback control laws for single-input single-output nonlinear processes with time delay are generalized to multivariable nonlinear processes. The controllers are based on two conceptually different approaches; the differential geometric approach is used for unconstrained nonlinear processes with deadtime in the controlled outputs and the model predictive approach is used for the constrained processes of the same type. The theoretical connections between model predictive control and input-output linearizing control are also established. A stirred tank reactor example is chosen to illustrate the performance of the proposed nonlinear controller in the presence of active input constraints.