OUTPUT PREDICTION UNDER RANDOM MEASUREMENTS. AN LMI APPROACH

In this paper, the design of an output predictor in a system with random scarce sampling is addressed. A model based predictor that takes into account the past measured outputs is used, and a Lyapunov function of the estimation error is used for design purposes. The Lyapunov design problem becomes a feasibility problem over a set of linear matrix inequalities applying the Schur complement formula. Three different design approaches have been developed. Some examples show the performances of each approach.