Comments on "Two-stage fractional least mean square identification algorithm for parameter estimation of CARMA systems"
暂无分享,去创建一个
This note points out some errors in the identification algorithms presented by M. A. Z. Raja and N. I. Chaudhary for estimation of controlled autoregressive moving average systems "Two-stage fractional least mean square identification algorithm for parameter estimation of CARMA systems", Signal Processing 107 (2015), 327-339.. Corrections on the basis of established techniques are also mentioned that solve the problems of the above mentioned paper. In addition, lack of any technical innovation between single stage and two-stage algorithms is highlighted. No technical difference between single stage and two-stage algorithms of 1.Noise v ( t ) is treated as known signal. This is totally wrong.Solutions of issues are presented on the basis of established results.
[1] Feng Ding,et al. Identification of Hammerstein nonlinear ARMAX systems , 2005, Autom..
[2] Raja Muhammad Asif Zahoor,et al. Two-stage fractional least mean square identification algorithm for parameter estimation of CARMA systems , 2015, Signal Process..
[3] Feng Ding,et al. Gradient-Based Identification Methods for Hammerstein Nonlinear ARMAX Models , 2006 .