INPUT-OUTPUT LINEARIZING NONLINEAR MODEL-PREDICTIVE CONTROL

Continuous-time nonlinear feedback control of multivariable processes with dead-times is considered. The problem of deadtime compensation fo r a class of nonlinear multivariable processes is solved. The nonlinea r controllers are derived by using two conceptually different approach es, a differential geometric approach and a model predictive approach. The model predictive approach uses shortest prediction horizons which allow one to establish the connections between model predictive contr ol and input-output linearization. These two approaches lead to identi cal feedback controllers when there are no constraints or when the cha racteristic (decoupling) matrix of process can be made diagonal by man ipulated-input/controlled-output rearrangements. Connections between t he derived controllers and the linear internal model control are estab lished. The application and performance of one of the derived nonlinea r controllers is illustrated by a reactor example.