Receding Horizon Filtering for Multisensor Linear Dynamics Systems

Distributed receding horizon discrete-time filtering is presented here, which combines a Kalman filter and receding horizon strategy. Distributed fusion with the weighted sum structure is then applied to local receding horizon Kalman filters (LRHKFs) having non-equal horizon time intervals. The proposed distributed algorithm has a parallel structure that allows for the parallel processing of observations, thereby making it more reliable than the centralized version if some sensors become faulty. Moreover, the choice of receding horizon strategy makes the proposed algorithm robust against dynamic model uncertainties. Note that the derivation of the error cross-covariances between the LRHKFs is the key contribution in this distributed algorithm. The subsequent application of the proposed distributed filter to linear discrete-time dynamic systems within a multisensor environment demonstrates and confirms its effectiveness.