H/sub ∞/ Gaussian filter on infinite time horizon

Abstract

Addresses design of H/sub /spl infin// Gaussian filters on infinite time horizon. In particular, we are interested in the robust filter design in a Gaussian noisy environment for a target plant which is subject to model uncertainty. The proposed filter can be treated as a natural extension of the Kalman filter for plants with model uncertainty in their worst cases. It is shown that the design can be done by solving a set of coupled algebraic Riccati equations. It is claimed that the resulting filter captures a trade-off between H/sub /spl infin// and H/sub 2/ (Kalman) filters.