The performance of the normal Kalman filter (KF) will decrease when it is c
onstructed on the basis of an erroneous model. Xia rt al. (1994) proposed a
new adaptive state estimation algorithm to solve the divergence problem fo
r the KF, namely adaptive fading Kalman filter (AFKF). This paper discusses
the error covariance equation in the AFKF suggested by Xia er al. (1994).
(C) 1998 Elsevier Science Ltd. All rights reserved.