Synthesis for Linear Parameter Varying Systems

Saturday, May 31, 2014 - 2:00pm - 3:30pm
Lind 305
Peter Seiler (University of Minnesota, Twin Cities)
This talk reviews methods for synthesizing and analyzing controllers for
linear parameter varying (LPV) systems. This class of systems was widely studied
in the early 90's to formalize the use of ad-hoc gain-scheduling in
industry. The talk reviews the existing results for LPV systems and covers
recent progress on numerical algorithms for these systems. First, techniques
are described
to model practical systems within the LPV framework. Next, the existing
analysis and controller synthesis conditions are summarized. These conditions
lead, in some cases, to parameterized linear matrix inequalities (LMIs). The
talk discusses heuristic algorithms to approximately solve these parameterized
LMIs. Finally, recent results on robustness analysis for LPV systems using
integral quadratic constraints are provided.
MSC Code: