Parameterized LMIs in Control Theory

From MaRDI portal





The authors consider several control and optimization problems which can be transformed into parametrized linear matrix inequalities. They are as follows: quadratic programming including some nonconvex and nonconcave topics which can be converted into a Boolean programming task, Lyapunov function based stability and performance analysis for uncertain systems, structured singular values analyses for a class of matrix perturbation problems and linear parameter-varying control analysis and design. Using techniques based on a directional convexity concept, the authors are able to develop systematic relaxation techniques to convert the parametrized linear matrix inequalities into a standard linear matrix inequality problem. To illustrate the efficiency of the proposed approach and to compare it with other techniques the authors present some numerical examples resulting from robust stability analysis and linear parameter-varying synthesis problems. The main conclusion is that the technique is superior to the other approaches especially for problems with significantly more states than parameters, which is typical in control problems. Unfortunately this result is based only on experimental tests and has no theoretical proof.




Cited in
(47)








This page was built for publication: Parameterized LMIs in Control Theory

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4507407)