Síntese de observadores constantes para sistemas discretos com parâmetros variantes no tempo sujeitos a entradas desconhecidas e sua aplicação no projeto de controladores por realimentação de estados

Abstract

This work proposes synthesis conditions of constant observers for discrete-time linear parameter-varying systems subject to unknown inputs, as well as their application to the design of state-feedback controllers. Such approach does not require measuring or estimating the time-varying parameters of the systems, which is a recurrent premise in the literature. This also allows the proposed methodology to be applied to other system classes. An observer-based state-feedback control law with constant gain is proposed using the states estimated by the designed observer. The procedure of designing the observer-based controllers is rewritten in a structure similar to the static output-feedback problem, allowing techniques from the literature to be applied to design such controllers. The observer and the observer-based controller design conditions are described in terms of optimization problems subject to linear matrix inequalities constraints. Numerical examples are used to validate and evaluate the proposed approaches.