Optimal robust filtering for systems subject to uncertainties

In this paper we deal with an optimal filtering problem for uncertain discrete-time systems. Parametric uncertainties of the underlying model are assumed to be norm bounded. We propose an approach based on regularization and penalty function to solve this problem. The optimal robust filter with the respective recursive Riccati equation is written through unified frameworks defined in terms of matrix blocks. These frameworks do not depend on any auxiliary parameters to be tuned. Simulation results show the effectiveness of the robust filter proposed.