Podrobné informácie o publikácii

Autor(i):
M. Kvasnica – P. Bakaráč – M. Klaučo
Názov:
Complexity reduction in explicit MPC: A reachability approach
Časopis:
Systems & Control Letters
Rok:
2019
Kľúčové slovo(á):
Model predictive control, Complexity, Reachability, Parametric optimization, Mixed-integer optimization
Zväzok:
124
Strany:
19–26
Jazyk:
angličtina
Anotácia:
We propose to reduce the complexity of explicit MPC controllers by removing regions that will never be reached during the closed-loop evolution from a given set of initial conditions. The identification of such regions is done by solving a reachability analysis problem, formulated as a mixed-integer feasibility program. The procedure directly accounts for possible discrepancies between the prediction model and the actual plant dynamics by, among other things, considering a case where state measurements are affected by an unknown, but bounded measurement noise. The result of the procedure is the reduction of explicit MPC complexity without sacrificing closed-loop performance.
ISSN:
0167-6911
DOI:
10.1016/j.sysconle.2018.12.002
Kategória publikácie:
ADC – Vedecké práce v zahraničných karentovaných časopisoch
V3 – Vedecký výstup publikačnej činnosti z časopisu
Oddelenie:
OIaRP
Vložil/Upravil:
doc. Ing. MSc. Martin Klaučo, PhD.
Posledná úprava:
21.12.2018 12:48:41
Plný text:
Plný text nie je prístupný verejnosti, ale môžete požiadať o kópiu niektorého z autorov
BibTeX:
@article{uiam1980,
author={M. Kvasnica and P. Bakar\'a\v{c} and M. Klau\v{c}o},
title={Complexity reduction in explicit MPC: A reachability approach},
journal={Systems \& Control Letters},
year={2019},
keyword={Model predictive control, Complexity, Reachability, Parametric optimization, Mixed-integer optimization},
volume={124},
pages={19-26},
annote={We propose to reduce the complexity of explicit MPC controllers by removing regions that will never be reached during the closed-loop evolution from a given set of initial conditions. The identification of such regions is done by solving a reachability analysis problem, formulated as a mixed-integer feasibility program. The procedure directly accounts for possible discrepancies between the prediction model and the actual plant dynamics by, among other things, considering a case where state measurements are affected by an unknown, but bounded measurement noise. The result of the procedure is the reduction of explicit MPC complexity without sacrificing closed-loop performance.},
doi={10.1016/j.sysconle.2018.12.002},
url={https://www.uiam.sk/assets/publication_info.php?id_pub=1980}
}