Based on steady-state Kalman filter and white noise estimators, and according to thepole-assignment principle of the control theory, the pole-assignment fixed-interval steady-stateKalman smoother and Wiener smoother a...Based on steady-state Kalman filter and white noise estimators, and according to thepole-assignment principle of the control theory, the pole-assignment fixed-interval steady-stateKalman smoother and Wiener smoother are presented. They avoid computation of the initial opti-mal smoothing estimates and can rapidly eliminate the effects of the initial smoothing estimates byassigning the poles of the smoothers, so that they have a practical stability in the finite fixed inter-val. A simulation example shows their effectiveness.展开更多
For the discrete stochastic control systems with correlated noises, using the Kalman fil-tering method, based on the CARMA innovation model, a unified optimal fixed-interval whitenoise recursive Wiener smoother is der...For the discrete stochastic control systems with correlated noises, using the Kalman fil-tering method, based on the CARMA innovation model, a unified optimal fixed-interval whitenoise recursive Wiener smoother is derived. It contains high degree polynomial matrices with coef-ficient matrices exponentially decaying to zero. Further, by the truncation method, the corre-sponding fast suboptimal fixed-interval white noise Wiener smoothing algorithm is presented,which obviously reduces the computational burden. The error formula of the smoother and theformula of selecting the truncated index are given. A simulation example for Bernoulli-Gaussianwhite noise shows the effectiveness of the proposed results.展开更多
文摘Based on steady-state Kalman filter and white noise estimators, and according to thepole-assignment principle of the control theory, the pole-assignment fixed-interval steady-stateKalman smoother and Wiener smoother are presented. They avoid computation of the initial opti-mal smoothing estimates and can rapidly eliminate the effects of the initial smoothing estimates byassigning the poles of the smoothers, so that they have a practical stability in the finite fixed inter-val. A simulation example shows their effectiveness.
文摘For the discrete stochastic control systems with correlated noises, using the Kalman fil-tering method, based on the CARMA innovation model, a unified optimal fixed-interval whitenoise recursive Wiener smoother is derived. It contains high degree polynomial matrices with coef-ficient matrices exponentially decaying to zero. Further, by the truncation method, the corre-sponding fast suboptimal fixed-interval white noise Wiener smoothing algorithm is presented,which obviously reduces the computational burden. The error formula of the smoother and theformula of selecting the truncated index are given. A simulation example for Bernoulli-Gaussianwhite noise shows the effectiveness of the proposed results.
