ABSTRACT
The Kalman filtering is a very popular approach for estimating the states of a nominal system by using past measurements due to its simplicity, optimality, tractability, and robustness. A great number of results on the Kalman filter have been reported, and different approaches have been proposed. Noting that works on the filter designs, an implicit assumption is that the filter will be implemented exactly. The chapter is concerned with the problem of robust non-fragil Kalman filter design for linear systems with norm-bounded uncertainties. Such robust non-fragile Kalman filters are required to be robust with respect to uncertainties in both the plant and the Kalman filter gains. The chapter considers two classes of gain uncertainties, namely, additive and multiplicative, and design methods are given for such robust non-fragile Kalman filters in terms of solutions to algebraic Riccati equations.
