Title :
Recursive solutions of estimating equations and adaptive spectral factorization
Author :
Lai, Tze Leung ; Ying, Zhiliang
Author_Institution :
Dept. of Stat., Stanford Univ., CA, USA
fDate :
2/1/1992 12:00:00 AM
Abstract :
A recursive spectral factorization algorithm was developed by V. Solo (ibid., vol.34, no.10, pp.1047-1051, Oct. 1989) to be used with recursive instrumental variables for consistent estimation of the parameters of an ARMAX system, but the convergence of the spectral factorization algorithm is questionable. Herein, a modified adaptive spectral factorization algorithm is shown to converge and a general method for constructing convergent recursive solutions of nonlinear equations that are used to define parameter estimates is also presented
Keywords :
convergence; iterative methods; parameter estimation; ARMAX system; adaptive spectral factorization; convergent recursive solutions; iterative methods; nonlinear equations; parameter estimation; Adaptive equalizers; Linear matrix inequalities; Notice of Violation; Parameter estimation; Recursive estimation; Riccati equations; Statistics; Sufficient conditions; Symmetric matrices; Time varying systems;
Journal_Title :
Automatic Control, IEEE Transactions on
