Distributed Fusion Estimator With Time-dependent Multiplicative Noise System

Time-correlated multiplicative noises are widely existed in networked system-s and mobile communication systems.Based on projective theory and optimal weighted fusion algorithm in the linear minimum variance sense,this paper studies the distributed fusion estimation problem for multi-sensor stochastic systems with time-correlated multiplicative noise.The main contents are as follows:Formulti-sensor stochastic systems with first-order Gaussian Markov time-correlated multiplicative noises,the virtual states and virtual process noises are introduced to construct the recursive state equations of the virtual states.Based on an innovation analysis method,the local one-step predictors of system state and virtual state are designed.Then the local linear filter,multi-step predictor and smoother are designed based on the one-step predictor.The cross-covariance matrices between any two local estimation errors are derived.And then based on the matrix weighted,diagonal matrix weighted and scalar weighted fusion algorithms in the linear minimum variance sense,the corresponding distributed fusion estimators are given.Further,for the above systems,we consider that the additive noise is one-step auto-correlated and cross-correlated.By introducing virtual states,based on the state augmentation method,the original system is transformed into a stochastic system only with one-step auto-correlated and cross-correlated additive noises.At the same time,we consider the possible random packet dropouts over the network transmission.The measurement predictor is designed to compensate the random lost measurement data.For local single sensor system,the optimal linear filter in the linear minimum variance sense is proposed.The cross-covariance mtrices between any two local filtering errors are derived.Moreover,for multi-sensor system,the distributed fusion filter is given.