A Kalman filter algorithm using a moving window with applications
Wang Zheng-ou, JTANPING ZHANG · International Journal of Systems Science · 1995
One of the conditions of using a Kalman filter is to have a precise model. If the precise model does not exist, the precision of the filter will not be guaranteed, and the filter may even diverge. Unfortunately, it is difficult to obtain a precise model. Usually the model is obtained through estimation, therefore due to the accuracy of this estimation, there always exists a difference between the estimated model and the precise model. In order to prevent the filter from divergence and enhance the precision of the filter, a Kalman filter algorithm using a moving window (KFMW) is derived, and some applications of the KFMW are given in this paper.