[Home]   [  News]   [  Events]   [  People]   [  Research]   [  Education]   [Visitor Info]   [UCSD Only]   [Admin]
Home > Events > CCoM > Abstract
Search this site:

Randolph E. Bank
Philip E. Gill
Michael Holst

Administrative Contact:
Terry Le

Office: AP&M 7431
Phone: (858)534-9813
Fax: (858)534-5273
E-mail: tele@ucsd.edu
Measuring Degree of Controllability of a Linear Dynamical System

Emre Mengi
UCSD Department of Mathematics


A linear time-invariant dynamical system is controllable if its trajectory can be adjustedto pass through any pair of points by the proper selection of an input.Controllability canbe equivalently characterized as a rank problem and therefore cannot be verifiedreliably numerically in finite precision. To measure the degree of controllability of a systemthe distance to uncontrollability is introduced as the spectral orFrobenius norm of thesmallest perturbation yielding an uncontrollable system. For a first order system we present a polynomial time algorithm to find the nearest uncontrollable systemthat improves the computational costs of the previous techniques. The algorithm locates the globalminimum of a singular value optimization problem equivalent to the distance to uncontrollability.In the second part for higher-order systems we derive a singular-valuecharacterization and exploitthis characterization for the computation of the higher-order distance to+uncontrollability to lowprecision. Keywords: dynamical system, controllability, distance touncontrollability, Arnoldi, inverse iteration, eigenvalue optimization,polynomial eigenvalue problem

Tuesday, November 21, 2006
11:00AM AP&M 2402