Electronic Resource
Springer
Journal of optimization theory and applications
81 (1994), S. 277-296
ISSN:
1573-2878
Keywords:
Robust control
;
min-max techniques
;
synthesis methods
;
stability
;
multivariable control systems
Source:
Springer Online Journal Archives 1860-2000
Topics:
Mathematics
Notes:
Abstract An important (some say, the major) reason for using feedback control is the presence of uncertain parameters which are a natural part of any real dynamical model. In this paper, we consider uncertain constant parameters in a time-invariant linear plant and announce some new results concerning robust compensator synthesis. Using the min-max principle, we derive necessary conditions for fixed-order linear robust controllers assuring asymptotic stability or relative stability. These necessary conditions are an extension of the Lagrange multiplier method. This is achieved using a cost function based on the inverse of the so-called critical constraint. We present both matrix and polynomial versions; the latter allows controllers of fixed structure. We suggest a probability-one homotopy algorithm and solve some examples from the literature.
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1007/BF02191665
|
Location |
Call Number |
Expected |
Availability |
