Finite-dimensional adaptive observers applied to distributed parameter systems

Author

Lilly, John H.

Author_Institution

Dept. of Electr. Eng., Louisville Univ., KY, USA

fYear

1990

fDate

5-7 Dec 1990

Firstpage

3558

Abstract

The application of adaptive observers to distributed parameter systems is considered. The effects of the infinite-dimensional unmodeled dynamics (residuals) on the observer state and parameter estimates are analyzed. For the purpose of the analysis, the state and parameter estimates of the nth-order observer are interpreted as estimates of an nth-order reduced-order model (ROM) of the infinite-dimensional plant. These estimates can then be compared to the states and parameters of a true ROM and bounds for the errors can be derived. It is proven that if the input provides definite excitation and the residual energy is bounded over a finite time interval, the estimation errors are ultimately bounded, with no assumption regarding stability of the plant or boundedness of the plant input

Keywords

State estimation; adaptive systems; distributed parameter systems; multidimensional systems; parameter estimation; state estimation; distributed parameter systems; finite dimensional adaptive observers; infinite-dimensional unmodeled dynamics; reduced-order model; residuals; Boundary conditions; Distributed parameter systems; Hilbert space; Observers; Parameter estimation; Read only memory; Reduced order systems; Sensor systems; Stability; State estimation;

fLanguage

English

Publisher

ieee

Conference_Titel

Decision and Control, 1990., Proceedings of the 29th IEEE Conference on

Conference_Location

Honolulu, HI

Type

conf

DOI

10.1109/CDC.1990.203487

Filename

203487