Distributed Kalman filter with fuzzy noises over multiagent systems under switching topology

In distributed state estimation, there exist several systems with fuzzy processes and observation noises, and adherence to a Gaussian distribution is insufficient; thus, the probability assumption no longer remains appropriate. To study this problem, we model the noises as fuzzy random variables with trapezoidal probability distributions using four representative points instead of Gaussian distributions. The state estimations of different nodes are then fused using fuzzy information fusion based on consensus. Furthermore, because the network is not always fixed in general scenarios, the changed network is modeled as a switching topology model. Subsequently, a distributed fuzzy Kalman filter algorithm under switching topology is proposed and the algorithm stability is analyzed. Finally, we demonstrate the effectiveness of the proposed estimation algorithm by applying it to a target-tracking problem.