Robust \(H_{\infty}\) filtering with error variance constraints for discrete time-varying systems with uncertainty (Q1398392)

The problem of robust \(H_\infty\)-filtering over a finite horizon is considered for discrete time-varying systems subject to unmeasured parameter perturbations in both the state and the output matrices of the state-space model. Sufficient conditions for the finite-horizon filter to satisfy state estimation error variance constraints as well as a prescribed \(H_\infty\)-performance for all admissible perturbations are given in terms of two unilaterally coupled discrete Riccati difference equations. The Riccati difference equations can be implemented for recursive computation in online applications. The results for time-varying systems has been extended to cover the case of stationary filtering over an infinite horizon for time-invariant systems. An example for estimating an inaccessible output of a time-varying uncertain system has been provided to illustrate how the special free parameter can be exploited to ensure that the bounds on the state estimation error variances stay below prescribed levels.