Time - Varying Systems Multi-sensor Information Fusion Filter And Its Application

For discrete time-varying linear stochastic system with multi-sensor and with correlated noises, using Kalman filtering method, based on the Riccati equation, under the linear minimum variance optimal information fusion criterion, three distributed optimal information fusion Kalman estimators weighted by matrices, diagonal matrices and scalars, are respectively presented, which can handle the fused filtering, prediction and smoothing problems in a unified framework. Three distributed optimal information fusion white noise deconvolution filters and smoothers are also presented respectively. The formulas and equivalent algorithms of computing the local estimation error covariances are presented, which are used in computing the optimal weighted. In addition, under the optimal weighted measurement fusion criterion, two weighted measurement fusion global optimal white noise deconvolution filters and smoothers are presented. The Monte-Carlo simulation examples in tracking systems and the numerical Monte-Carlo simulation examples show their effectiveness.