Model reference control for a nonlinear production-inventory system

TOSETTI SANTIAGO; PATIÑO DANIEL; CAPRARO FLAVIO; GAMBIER ADRIAN

Rosario

Congreso; 37º JAIIO - Jornada de Informática Industrial 2008, Cadena de Abastecimiento; 2008

Sociedad Argentina de Informática

In this paper an approach to model reference control is proposed and analyzed for a simple class of production-inventory system. In this case the production system is modeled as a first order process with a nonlinear output. The inventory has size restrictions which are also modeled by a nonlinear saturation function. The controller uses a second order nonlinear system as model reference, and the controller structure and parameters are found by analyzing the error dynamical equations. The controller needs also a prediction of the demand that is obtained from a Dual Extended Kalman Filter. In this approach the Kalman filter estimates not only the process states but also the parameters of a Volterra model, proposed for modeling the demand. Simulation studies are finally presented in order to show the feasibility and performance of the proposed controller.