Off-line estimation of dynamic systems parameters in the presence of outliers in observations

Article type: Research Article

Authors: Pupeikis, Rimantas

Affiliations: Institute of Mathematics and Informatics, 2600 Vilnius, Akademijos St. 4, Lithuania

Abstract: In the previous papers (Pupeikis, 1990; 1991; 1992) the problems of model oder determination and recursive estimation of dynamic systems parameters in the presence of outliers in observations have been considered. The aim of the given paper is the development, in such a case, of classical off-line algorithms for systems of unknown parameters estimation using batch processing of the stored data. An approach, based on a substitution of the corresponding values of the sample covariance and cross-covariance functions by their robust analogues in respective matrices and on a further application of the least square (LS) parameter estimation algorithm, is worked out. The results of numerical simulation by IBM PC/AT (Table 1, 2) are given.

Keywords: LS algorithm, covariance analysis, outlier, robustness

DOI: 10.3233/INF-1992-3407

Journal: Informatica, vol. 3, no. 4, pp. 567-581, 1992