Pole-assignment fixed-interval Kalman smoother with an exponential stability |
| |
Authors: | S-L Sun Z-L Deng |
| |
Affiliation: | 1. Department of Automation , Heilongjiang University , P.O. Box 130, Harbin 150080, People's Republic of China sunsl@hlju.edu.cn sunsl216@tom.com;3. Department of Automation , Heilongjiang University , P.O. Box 130, Harbin 150080, People's Republic of China |
| |
Abstract: | Using the innovation analysis method in the time domain, based on the autoregressive moving average (ARMA) innovation model and white noise estimators, a pole-assignment fixed-interval steady-state Kalman smoother is presented for discrete-time linear stochastic systems. It avoids the computation of the optimal initial smoothing estimate, and can rapidly eliminate the effect of arbitrary initial smoothing estimate by assigning the poles of the smoother, with an exponentially decaying rate. Several simulation examples show its effectiveness. |
| |
Keywords: | State estimation Fixed-interval Kalman smoother Pole-assignment ARMA innovation model Exponential stability |
|
|