A novel adaptive Kalman filtering algorithm for ARMA signal
Guo Dianlong, Hexin Chen, Dai Yisong · 2002
A kind of adaptive Kalman filtering algorithm suited to autoregressive moving average (ARMA) signals is proposed. If the parameters of the signal and the noise variances are unknown, a three-stage recursive least squares approach is first adopted to estimate the parameters of the corrupted signal, then the scalar estimate presented is used to estimate the output value of filtering based on the estimated parameters and the current observation value. Use of the algorithm simplifies and speeds up the filtering process because the regressive matrix equations need not be solved. The adaptive Kalman filtering algorithm can be used in the form of a series connection to improve the accuracy of parameter recognition. Computer simulation experiments show that the algorithm is effective.>