Round-off error propagation in four generally-applicable, recursive, least-squares estimation schemes

Automatica

Volumn 25, Issue 3, 1989, Pages 437-444

  Verhaegen, M H ^a  

^a DELFT UNIVERSITY OF TECHNOLOGY   (Netherlands)

Author keywords

computer simulation; least squares estimation, numerical analysis; Recursive algorithms

Indexed keywords

COMPUTER SIMULATION; MATHEMATICAL TECHNIQUES--LEAST SQUARES APPROXIMATIONS; SIGNAL FILTERING AND PREDICTION--KALMAN FILTERING; STATISTICAL METHODS--REGRESSION ANALYSIS;

KALMAN FILTER; LINEAR REGRESSION MODEL; RECURSIVE LEAST SQUARES; ROUND-OFF ERRORS;

CONTROL SYSTEMS;

EID: 0024666768     PISSN: 00051098     EISSN: None     Source Type: Journal    
DOI: 10.1016/0005-1098(89)90013-7     Document Type: Article

References (15)

1. Theory and applications of self-tuning regulators
2. The numerical stability of the Levison-Durbin algorithm for Toeplitz systems of equations
3. Methods for modifying matrix factorizations
4. New estimation techniques for adaptive control
5. Turn-off, escape and probing in non-linear stochastic control
6. Restricted exponential forgetting in real-time identification
7. The Weiner RMS (root mean square) error criterion in filter design and prediction
8. Error propagation properties of recursive least squares adaptation algorithms
9. Stability of the solutions of linear least squares problems
10. Perturbation bounds for the QR factorization of a matrix
11. A new class of algorithms in linear system theory, with application to real-time aerodynamic model identification
12. Numerical aspects of different Kalman filter implementations