Reduced-order suboptimal filter for dynamic systems with multi-sensor environment

This paper considers the problem of fusion of local Kalman filters for dynamic systems with multi-sensor environment. The filter performance in multi-sensor dynamic system is effected because of communication restriction, data association and estimation errors. A new reduced-order suboptimal filter is proposed for multi-sensor dynamic systems which reduces the computational cost for state estimation. The filtering algorithm includes two stages: the locally optimal Kalman estimates computed at the first stage are linearly fused at the second stage. The proposed filter has parallel structure and is suitable for parallel processing of measurements which can also help to minimize the computation time and produce real time state estimation. Example of systems containing different types of sensors, demonstrating the accuracy of the proposed filter, are given.