Robust integrated covariance intersection fusion Kalman estimators for networked systems with a unified measurement model including five uncertainties

Abstract

For networked multisensor systems (NMSs), a soft measurement model with five uncertainties is presented, which can be viewed as a “soft sensor” such that the measurements received by estimators can be obtained via the computations of this model. A novel “soft sensor” concept is presented. Three new approaches are presented, which include a fictitious white noises approach with compensating random uncertainties, an extended Lyapunov equation approach with three kinds of the Lyapunov equations, and a universal integrated covariance intersection (ICI) fusion approach with integrating local estimators and their cross-covariances. Applying them, the universal ICI and two fast ICI (FICI) fusion time-varying and steady-state minimax robust Kalman estimators (predictor, filter and smoother) are presented in the sense that their actual estimation error variances are guaranteed to have the corresponding minimal upper bounds. They improve the robust accuracies of the original covariance intersection (CI) and fast CI(FCI) fusers, and overcome their drawback to require known local estimators and their conservative variances. Their robustness, accuracy relations, stability, steady-state property, and three modes of convergence are proved. The proposed new concept, approaches and results constitute a new methodology, a universal robust fusion Kalman filtering theory and a new filtering stability theory. A simulation example applied to the vehicle suspension system shows their effectiveness and applicability.