DYNAMIC OUTPUT-FEEDBACK CONTROL OF MINIMUM-PHASE MULTIVARIABLE NONLINEAR PROCESSES

This paper concerns the synthesis of dynamic output feedback controlle rs for minimum-phase multivariable nonlinear processes with a nonsingu lar characteristic matrix. State-space controller realizations are der ived that induce a linear input/output behavior of general form in the closed-loop system. A combination of input/output linearizing state f eedback laws and state observers is employed for the derivation of the controllers. For open-loop stable processes, the process model is use d as an open-loop state observer. In the more general case of possible open-loop instability, a reduced-order observer is used based on the forced zero dynamics of the process model. The performance and robustn ess characteristics of the proposed control methodology are illustrate d through simulations in a chemical reactor example.