Sequential Fast Covariance Intersection Fusion Kalman Filter for Networked System with Fading Measurements and Correlated Noises

For the networked systems with fading measurements, multiplicative noises and correlated noises, the model transformation technique with the fictitious noises and Sequential Fast Covariance Intersection (SFCI) fusion Kalman filter are presented. These algorithms can compensate the mixed random uncertainties with the fictitious noises and decrease the large computational burden. Furthermore, the SFCI fusion algorithm overcomes the drawback of the Sequential Covariance Intersection fusion, whose fusion result is sensitive to the fusion orders, and improves the accuracy and efficiency. The accuracy of the presented SFCI fusion Kalman filter is higher than that of each local estimator, less but close to that of the information fusion Kalman filter weighted by matrices. Finally, the effectiveness of the SFCI fusion Kalman algorithm is verified by the simulation example.