@inproceedings{Scarciotti:2016:10.1109/CDC.2016.7799420,
author = {Scarciotti, G and Astolfi, A and Jiang, Z-P},
doi = {10.1109/CDC.2016.7799420},
publisher = {IEEE},
title = {Constrained optimal reduced-order models from input/output data},
url = {http://dx.doi.org/10.1109/CDC.2016.7799420},
year = {2016}
}

Download

TY  - CPAPER
AB  - Model reduction by moment matching does notpreserve, in a systematic way, the transient response of thesystem to be reduced, thus limiting the use of this modelreduction technique in control problems. With the final goalof designing reduced-order models which can effectively beused (not just for analysis but also) for control purposes, wedetermine, using a data-driven approach, an estimate of themoments and of the transient response of an unknown system.We compute the unique, up to a change of coordinates, reducedordermodel which possesses the estimated transient and,simultaneously, achieves moment matching at the prescribedinterpolation points. The error between the output of the systemand the output of the reduced-order model is minimized andwe show that the resulting system is a constrained optimal (ina sense to be specified) reduced-order model. The results of thepaper are illustrated by means of a simple numerical example.
AU  - Scarciotti,G
AU  - Astolfi,A
AU  - Jiang,Z-P
DO  - 10.1109/CDC.2016.7799420
PB  - IEEE
PY  - 2016///
TI  - Constrained optimal reduced-order models from input/output data
UR  - http://dx.doi.org/10.1109/CDC.2016.7799420
UR  - http://hdl.handle.net/10044/1/38599
ER  - 

Download