DocumentCode
2476124
Title
Algebraically Defined Gramians for Nonlinear Systems
Author
Gray, W. Steven ; Verriest, Erik I.
Author_Institution
Dept. of Electr. & Comput. Eng., Old Dominion Univ., Norfolk, VA
fYear
2006
fDate
13-15 Dec. 2006
Firstpage
3730
Lastpage
3735
Abstract
In this paper, controllability and observability Gramian definitions central to the model reduction method for bilinear state space systems due to Al-Baiyat, et al. are extended to the nonlinear affine control case. While in general these algebraically defined Gramians lack a direct physical interpretation, their numerical tractability is superior to the established general nonlinear Gramian generalizations known as the energy functions. The latter can only be determined at present by either solving a pair of Hamilton-Jacobi-Bellman equations or by Monte-Carlo simulation. Connections between the algebraic Gramians developed here and existing nonlinear Gramian generalizations are presented. A connection to approximation theory for nonlinear operators is also described. Cascade lemma for set-stable systems
Keywords
controllability; matrix algebra; nonlinear control systems; observability; Hamilton-Jacobi-Bellman equations; Monte Carlo simulation; algebraically defined Gramians; approximation theory; bilinear state space systems; controllability Gramian definition; energy functions; model reduction method; nonlinear Gramian generalizations; nonlinear affine control; nonlinear systems; numerical tractability; observability Gramian definition; Approximation methods; Centralized control; Controllability; Differential equations; Nonlinear control systems; Nonlinear equations; Nonlinear systems; Observability; Reduced order systems; State-space methods;
fLanguage
English
Publisher
ieee
Conference_Titel
Decision and Control, 2006 45th IEEE Conference on
Conference_Location
San Diego, CA
Print_ISBN
1-4244-0171-2
Type
conf
DOI
10.1109/CDC.2006.376840
Filename
4177638
Link To Document