Self-tuning Of Distributed Information Fusion State Estimators And Its Application

For the multisensor linear time-invariant discrete stochastic systems with unknown noise statistics, using the modern time series analysis method, based on the autoregressive moving average (ARMA) innovation model, and based on the solution of the matrix equations for correlation function, the estimators of the noise statistics are obtained. Under the optimal fusion rules weighted by matrices, diagonal matrices and scalars, three self-tuning information fusion Kalman estimators are presented respectively. The self-tuning decoupled information fusion Wiener estimators weighted by scalars and diagonal matrices are also presented for state components. They can handle the fused filtering, smoothing and prediction problems in an unified framework. Their convergence (asymptotic optimality) is proved, i.e. if the parameter estimation of ARMA innovation model is consistent, they will converge to the optimal information estimators in a realization. Many simulation examples for the target tracking systems show their effectiveness.