Adaptive internal model control of nonlinear processes

Model-based controllers are often essential for effective control of nonlin ear processes. Performance and robustness of these controllers are affected by the inevitable modeling errors, and parameter adaptation is a technique to robustify the model-based controllers. In this paper, an adaptive inter nal model control (AdIMC) for a class of minimum-phase input-output lineari zable nonlinear systems with parameter uncertainty is presented. Internal m odel control (IMC) for nonlinear systems is developed directly from input-o utput linearization. The parameter adaptation for the IMC is based on proce ss and model outputs, and the state variables predicted by the model only. Asymptotic tracking and convergence of unknown parameters by the proposed a daptation, is first shown theoretically. Then, AdIMC is applied to two nonl inear processes (a fermenter and a neutralization process), and its perform ance for a variety of disturbances and modeling errors is studied. The theo retical and simulation results show that the proposed AdIMC improves the pe rformance and robustness of the IMC controller for nonlinear processes. Als o, the proposed adaptation can easily be implemented in the IMC structure. (C) 1999 Elsevier Science Ltd. All rights reserved.