Distributed receding horizon filtering in discrete-time dynamic systems

A distributed receding horizon filtering for discrete-time dynamic systems is proposed. A distributed fusion with the weighted sum structure is applied to the set of local receding horizon Kalman filters (LRHKFs). All LRHKFs have the same receding horizon length. The distributed fusion algorithm represents the optimal linear fusion by weighting matrices under the minimum mean square criterion. In other to compute the optimal matrix weights, the recursive equations for error cross-covariances between the LRHKFs are derived. Simulation example for the tracking system with three sensors demonstrates effectiveness of the proposed filter.