Distributed Luenberger observers for linear systems

The distributed Luenberger observers for linear systems under switching topology is considered. These observers are arranged in a communication graph configuration, and perform estimation tasks by distributed local data fusion in the sense that each observer receives measurements from local sensors and exchanges information with its neighbors. The objective is convergence of each observer state to that of the given linear system. Leader-following consensus algorithms is applied to the distributed observers design. The communication graph allows to be time-varying. A modified averaging approach is utilized to aid the convergence analysis under the jointly connected graph topology. An illustrated example is presented to validate the result.