An Optimisation-based Approach to Interval Model Identification in the Frequency Domain
| Authors: | Puig Vicenç, Universitat Politècnica de Catalunya, Spain
Masip Albert, Universitat Politècnica de Catalunya, Spain
Quevedo Joseba, Universitat Politècnica de Catalunya, Spain |
|---|
| Topic: | 1.1 Modelling, Identification & Signal Processing |
|---|
| Session: | Frequency Domain Estimation Methods |
|---|
| Keywords: | Frequency Response, Identification, Uncertain Systems, Robustness |
|---|
Abstract
In this paper, an Optimisation-based Approach to Interval Model Identification in the Frequency Domain is proposed. As a result of its application, an interval LTI model with a prefixed structure that satisfies that its worst-case frequency response contains all the measured given data is obtained. The optimisation problem is formulated using, as constraints, that this inclusion property is explicitly satisfied making use of some auxiliary results used to prove the Kharitonov Theorem and, as objective function, the minimisation of the size of parameter space box according to some metric and the width of model frequency envelopes. Finally, an example of application of this approach is proposed.
