Talks and Poster Presentations (with Proceedings-Entry):
A. Schirrer, M. Kozek:
"Assessment of Interpolation Methods for Linear-Dynamic Systems";
Talk: CSSim2011, 2nd International Conference on Computer Modelling and Simulation,
Brno, Czech Republic;
2011-09-05
- 2011-09-07; in: "Proceedings of the 2nd International Conference on Computer Modelling and Simulation-CSSim2011",
(2011),
ISBN: 9788021443204;
137
- 145.
English abstract:
Local linear models are a versatile structure to model nonlinear
/ parameter-varying systems. However, a meaningful interpolation
between the individual models is not a trivial task. Given a set of
LTI state space models ("grid models") at known parameter grid points,
linear system interpolation aims to generate systems with meaningful
dynamics at parameter values between these parameter values. Three
interpolation methods are reviewed in this paper: companion form interpolation,
modal form interpolation, and a novel Geometric Algebra
(GA)-based interpolation. Stability and other key properties are assessed
and compared. A discussion outlines the major differences and areas of
applicability. The novel GA-based method provides an interesting, geometric
view on interpolation and is subject of future research.
Keywords:
modelling, interpolation, eigenstructure, stability
Created from the Publication Database of the Vienna University of Technology.