Modified robust covariance intersection fusion steady-state Kalman estimators with uncertain noise variances

For multisensor time-invariant system with uncertain noise variances. According to the minimax robust estimation principle, applying the non-augmented state approach and the direct approach, the unified modified robust CI fusion steady-state Kalman estimators (predictor, filter and smoother) are presented using the conservative cross-covariances of the local estimation errors. Their actual error variances are guaranteed to have the corresponding minimal upper-bound. Their robust accuracies are higher than that of the original CI fuser. A simulation example is given to verify the robustness and robust accuracy relations.