Summary: This paper deals with an adaptive control algorithm which consists of combining an input-output model reference control design with a robust parameter estimator with respect to external disturbances, time-varying parameters and unmodeled dynamics. The input-output reference model concept is basically due to a suitable interpretation of the pole placement principle. This allows to incorporate an adequate tracking capability into the design of those quadratic cost function minimization based control laws, as the generalized predictive control.
An applicability study, involving a liquid-liquid extraction column, is reported to emphasize the performances of the proposed adaptive controller.