New approach to Wiener state filtering in time-domain
时域Wiener状态滤波新方法

Based on the steady-state Kalman filter and projection theory,a new unified and general approach to the time-domain Wiener state filtering is presented,by which the asymptotically stable Wiener state estimator and decoupled Wiener state estimators are presented for linear stochastic systems with correlated noises having non-zero means.It can handle the state filtering,prediction and smoothing problems in a unified framework.The transformation relationship between the Kalman filters and Wiener filters is discovered,the Wiener state estimators can be obtained from the Kalman estimators by means of the autoregressive moving average (ARMA) innovation model.A simulation example shows its effectiveness.