DocumentCode
1431403
Title
A direct characterization of L2-gain controllers for LPV systems
Author
Kose, I.E. ; Jabbari, F. ; Schmitendorf, W.E.
Author_Institution
Dept. of Mech. & Aerosp. Eng., California Univ., Irvine, CA, USA
Volume
43
Issue
9
fYear
1998
fDate
9/1/1998 12:00:00 AM
Firstpage
1302
Lastpage
1307
Abstract
In this paper, a class of linear parameter dependent output feedback controllers that satisfy quadratic stability and an induced L 2-norm bound for a given linear parameter-varying (LPV) plant is considered. By using a parameter-independent common Lyapunov function, the solvability conditions are expressed in terms of finite dimensional linear matrix inequalities (LMI) evaluated at the extreme points of the admissible parameter set. Conditions under which strictly proper controllers can be used are obtained. By restricting some of the controller matrices to be constant, the input and output matrices can be parameter-varying without destroying the problem´s convexity. Cases where the controller matrices can be obtained without interpolation are also discussed, thereby simplifying the implementation of the controller. A numerical example is included which demonstrates the application of the results
Keywords
H∞ control; Lyapunov methods; feedback; linear systems; matrix algebra; stability; L2-gain controllers; LMI; LPV systems; convexity; finite dimensional linear matrix inequalities; induced L2-norm bound; input matrix; linear parameter dependent output feedback controllers; linear parameter-varying plant; output matrix; parameter-independent common Lyapunov function; quadratic stability; solvability conditions; strictly proper controllers; Circuits; Concurrent computing; Control systems; Discrete event systems; Linear feedback control systems; Linearity; Optimized production technology; Production systems; Stochastic processes; Throughput;
fLanguage
English
Journal_Title
Automatic Control, IEEE Transactions on
Publisher
ieee
ISSN
0018-9286
Type
jour
DOI
10.1109/9.718622
Filename
718622
Link To Document