Distributed Robust Estimation with Dynamics Uncertainties and Random Communication Topologies

This paper discusses the design of distributed structures and gains for estimation of sensor network with nonlinear uncertain dynamics. Each sensor estimates the system state not only on its own but also on its neighbors' information. Besides, these sensors communicate with their neighbors over a random communication topology, i.e., their measurements are transmitted randomly at every step. In addition, both the system uncertainties and its linearization errors are taken into consideration in this paper. It is strictly proved that an optimized upper bound of the estimation error covariance with respect to the proposed filtering method can be achieved.