Sequential Inverse Covariance Intersection Fusion Kalman Predictor for Multi-sensor Systems with Packet Dropouts and Correlated Noise

This paper focuses on the problem of the multi-sensor state estimation for the networked system with packet dropouts and linear correlated additive white noises. By the state augmented method and the fictitious noise technology, the original system is transformed into a standard system with white noises, and then applying the local Kalman predictor, the SICI fusion estimator is designed by the inverse covariance intersection fusion algorithm. Compared with the CI fusion algorithm, the SICI fusion algorithm presented in this paper has a smaller computational burden. Compared with the SCI fusion algorithm, the SICI fusion algorithm has a lower degree of conservativeness and provides more accurate fusion results. It is proved that the estimation accuracy of the SICI fusion algorithm is higher than that of SCI fusion, and the superiority of the algorithm is verified by a simulation example.