Accuracy analysis of bias-eliminating least squares estimates for errors-in-variables identification

Mei Hong; Torsten Söderström; Wei Xing Zheng

Accuracy analysis of bias-eliminating least squares estimates for errors-in-variables identification

Mei Hong
, Torsten Söderström
, Wei Xing Zheng

School of Computer, Data & Mathematical Sciences

Research output: Chapter in Book / Conference Paper › Conference Paper

Abstract

The bias-eliminating least squares (BELS) method is one of the consistent estimators for identifying dynamic errors-in-variables systems. The attraction of the BELS method lies in its good accuracy and its modest computational cost. In this paper, we investigate the accuracy properties of the BELS estimates. It is shown that the estimated system parameters and the estimated noise variances are asymptotically Gaussian distributed. An explicit expression for the normalized covariance matrix of the estimated parameters is derived and supported by some numerical examples.

Original language	English
Title of host publication	System Identification, Vol. 14, No. 1
Publisher	IFAC in partnership with Elsevier
Number of pages	6
Publication status	Published - 2007
Event	IFAC Symposium on System Identification - Duration: 11 Jul 2012 → …

Conference

Conference	IFAC Symposium on System Identification
Period	11/07/12 → …

Keywords

system identification
parameter estimation
errors-in-variables
bias-eliminating least squares
accuracy analysis

Cite this

@inproceedings{82725398a58a47f5819b0b0910dd6ab5,

title = "Accuracy analysis of bias-eliminating least squares estimates for errors-in-variables identification",

abstract = "The bias-eliminating least squares (BELS) method is one of the consistent estimators for identifying dynamic errors-in-variables systems. The attraction of the BELS method lies in its good accuracy and its modest computational cost. In this paper, we investigate the accuracy properties of the BELS estimates. It is shown that the estimated system parameters and the estimated noise variances are asymptotically Gaussian distributed. An explicit expression for the normalized covariance matrix of the estimated parameters is derived and supported by some numerical examples.",

keywords = "system identification, parameter estimation, errors-in-variables, bias-eliminating least squares, accuracy analysis",

author = "Mei Hong and Torsten S{\"o}derstr{\"o}m and Zheng, \{Wei Xing\}",

year = "2007",

language = "English",

booktitle = "System Identification, Vol. 14, No. 1",

publisher = "IFAC in partnership with Elsevier",

note = "IFAC Symposium on System Identification ; Conference date: 11-07-2012",

}

TY - GEN

T1 - Accuracy analysis of bias-eliminating least squares estimates for errors-in-variables identification

AU - Hong, Mei

AU - Söderström, Torsten

AU - Zheng, Wei Xing

PY - 2007

Y1 - 2007

N2 - The bias-eliminating least squares (BELS) method is one of the consistent estimators for identifying dynamic errors-in-variables systems. The attraction of the BELS method lies in its good accuracy and its modest computational cost. In this paper, we investigate the accuracy properties of the BELS estimates. It is shown that the estimated system parameters and the estimated noise variances are asymptotically Gaussian distributed. An explicit expression for the normalized covariance matrix of the estimated parameters is derived and supported by some numerical examples.

AB - The bias-eliminating least squares (BELS) method is one of the consistent estimators for identifying dynamic errors-in-variables systems. The attraction of the BELS method lies in its good accuracy and its modest computational cost. In this paper, we investigate the accuracy properties of the BELS estimates. It is shown that the estimated system parameters and the estimated noise variances are asymptotically Gaussian distributed. An explicit expression for the normalized covariance matrix of the estimated parameters is derived and supported by some numerical examples.

KW - system identification

KW - parameter estimation

KW - errors-in-variables

KW - bias-eliminating least squares

KW - accuracy analysis

UR - http://handle.uws.edu.au:8081/1959.7/36421

M3 - Conference Paper

BT - System Identification, Vol. 14, No. 1

PB - IFAC in partnership with Elsevier

T2 - IFAC Symposium on System Identification

Y2 - 11 July 2012

ER -

Accuracy analysis of bias-eliminating least squares estimates for errors-in-variables identification

Abstract

Conference

Keywords

Other files and links

Fingerprint

Cite this