Distributed Fusion Filter for Systems With Time-Correlated Multiplicative Noise

The distributed fusion state estimation problem for systems with time-correlated multiplicative noise is studied where the time-correlated multiplicative noise is described by a first-order Gauss-Markov process. Based on the projection theory, the distributed fusion filter weighted by scalars is given in the linear minimum variance (LMV) sense. The filtering error cross-covariance matrices between any two local filters are derived to obtain the fusion weights. The simulation research verifies the effectiveness of the proposed algorithms.