Performance bounds of discrete-time optimal FIR filter under modeling uncertainty

In this paper, estimation error bounds of the optimal FIR filter proposed by Kwon et al. (1989, 1990), are presented in discrete-time systems with model uncertainty. Performance bounds are represented by the upper bounds on the difference of the estimation error covariances between the nominal and real values in the case of systems with noise or model parameter uncertainty. The estimation error bounds of the discrete-time optimal FIR filter is compared with those of the Kalman filter via a numerical example applied to the simulation problem by Toda and Patel (1990). Simulation results show that the former has a more robust performance than the latter