On time and frequency characteristics of weighted least squares estimators applied to nonstationary system identification

作者: Maciej Niedzwiecki

DOI: 10.1109/CDC.1985.268814

关键词: Bandwidth (signal processing)Generalized least squaresNon-linear least squaresAlgorithmEstimatorSystem identificationImpulse responseMathematicsMathematical optimizationWeighting

摘要: After reviewing the already known "static" characteristics of weighted least squares (WLS) estimators such as effective and equivalent numbers observations, their tracking properties are investigated. Several new "dynamic" WLSE's proposed, like associate impulse response, frequency characteristics, estimation bandwidth delay. It is shown that may serve a convenient basis for comparison WLS corresponding to different weighting sequences, allow seldom raised question how fast parameter variations can be tracked successfully by this type (i.e. what WLSE's)to answered precisely.

参考文章(5)
M. Niedzwiecki, Bayesian-like autoregressive spectrum estimation in the case of unknown process order IEEE Transactions on Automatic Control. ,vol. 30, pp. 950- 961 ,(1985) , 10.1109/TAC.1985.1103818
Maciej Niedzwiecki, On the localized estimators and generalized Akaike's criterions 1981 20th IEEE Conference on Decision and Control including the Symposium on Adaptive Processes. ,vol. 29, pp. 56- 61 ,(1981) , 10.1109/CDC.1981.269442
J. Makhoul, L. Cosell, Adaptive lattice analysis of speech IEEE Transactions on Circuits and Systems. ,vol. 28, pp. 494- 499 ,(1981) , 10.1109/TCS.1981.1085022
J. Cioffi, T. Kailath, Fast, recursive-least-squares transversal filters for adaptive filtering IEEE Transactions on Acoustics, Speech, and Signal Processing. ,vol. 32, pp. 304- 337 ,(1984) , 10.1109/TASSP.1984.1164334