Robust guaranteed cost measurement fusion steady-state Kalman filters with uncertain noise variances

In this work, guaranteed cost robust centralized fusion (CF) and weighted measurement fusion (WMF) steady-state Kalman filters problems in multisensor system with uncertain noise variances are addressed. By parameterzing the perturbations of uncertain noise variances, two classes of guaranteed cost robust CF and WMF Kalman filters are proposed based on minimax robust estimation principle (MREP). The actual accuracy deviations are guaranteed to have the minimal upper bound and maximal lower bound. The proof of guaranteed cost robustness is proposed by using Lyapunov equation method. The equivalence of guaranteed cost WMF and CF Kalman filters is proved by using the information filter. A numerical example verifies the robustness of the proposed filters.