In vielen regelungstechnischen Anwendungen sind entweder das mathematische Systemmodell oder die Rahmenbedingungen, denen das geregelte System unterliegt, unbekannt oder unsicher. Dadurch sind klassische Regelungsmethoden oft nicht anwendbar. Datenbasierte Regelung beschäftigt sich daher mit dem Fall, dass kein Modell des Systems vorhanden ist und ermöglicht die Auslegung stabilisierender Regler anhand geeigneter Eingangs-/Ausgangs-Messungen des Systems. In ähnlicher Weise verwenden lernbasierte Regelungskonzepte künstliche Intelligenz oder lernende Algorithmen, um online entweder ein Modell, die Systemumgebung oder direkt einen Regler aus gemessenen Daten zu bestimmen. Diese Methoden haben ein großes Anwendungspotential in der Regelung nichtlinearer Systeme, bei denen die Ermittlung genauer Modelle oft aufwändig oder nicht praktikabel ist.
Unsere Forschung konzentriert sich insbesondere auf
- die Anwendung von Online Convex Optimization zur Regelung dynamischer Systeme,
- datenbasierte Beschreibungen von nichtlinearen Systemen,
- die Entwicklung datenbasierter prädiktiver Regelung,
- den Entwurf von off-policy Reinforcement Learning Algorithmen zur Ermittlung von Zustandsreglern
- die Analyse von datenbasierten Reglern für unregelmäßig abgetastete Systeme
- die Entwicklung von daten/lernbasierten Schätzern und
- die Anwendung systemtheoretischer Konzepte zur Entwicklung von lernbasierten Algorithmen mit theoretischen Garantien
Ausgewählte Veröffentlichungen
-
(2026): Online convex optimization for constrained control of nonlinear systems, Automatica, vol. 188, p. 112916
DOI: 10.1016/j.automatica.2026.112916
arXiv: 2412.00922 -
(2025): Online convex optimization for robust control of constrained dynamical systems, IEEE Transactions on Automatic Control (Early Access)
DOI: 10.1109/TAC.2025.3641731
arXiv: 2401.04487 -
(2025): Data-based system representations from irregularly measured data, IEEE Transactions on Automatic Control, vol. 70, no. 1, pp. 143-158
DOI: 10.1109/TAC.2024.3423053
arXiv: 2307.11589 -
(2025): Gaussian Process-Based Nonlinear Moving Horizon Estimation, IEEE Transactions on Automatic Control, Vol. 70, Iss. 12, pp. 7875 - 7890
DOI: 10.1109/TAC.2025.3580033
arXiv: 2402.04665 -
(2025): Notes on data-driven output-feedback control of linear MIMO systems, IEEE Transactions on Automatic Control, vol. 70, iss. 9, pp. 6143-6150
DOI: 10.1109/TAC.2025.3553073
arXiv: 2311.17484 -
(2024): An Input-Output Continuous-Time Version of Willems’ Lemma, IEEE Control Systems Letters, vol. 8, pp. 916-921
DOI: 10.1109/LCSYS.2024.3406057
arXiv: 2405.15482 -
(2024): Robust Data-Driven Moving Horizon Estimation for Linear Discrete-Time Systems, IEEE Transactions on Automatic Control, vol. 69, iss. 8, pp. 5598-5604
DOI: 10.1109/TAC.2024.3371373
arXiv: 2210.09017 -
(2023): Data-Based Control of Feedback Linearizable Systems, IEEE Transactions on Automatic Control, vol. 68, no. 11, pp. 7014-7021
DOI: 10.1109/TAC.2023.3249289
arXiv: 2204.01148 -
(2023): On the relation between dynamic regret and closed-loop stability, Systems & Control Letters, vol. 177, p. 105532
DOI: 10.1016/j.sysconle.2023.105532
arXiv: 2209.05964 -
(2023): Efficient Off-Policy Q-Learning for Data-Based Discrete-Time LQR Problems, IEEE Transactions on Automatic Control, vol. 68, no. 5, pp. 2922-2933
DOI: 10.1109/TAC.2023.3235967
arXiv: 2105.07761 -
(2021): Data-Driven Model Predictive Control With Stability and Robustness Guarantees, IEEE Transactions on Automatic Control, 2021, Vol. 66, No. 4, pp. 1702-1717
DOI: 10.1109/TAC.2020.3000182
Ausgewählte Projekte
-
Cont4Med - Estimation and control under limited information with application to biomedical systemsLeitung: Prof. Dr.-Ing. Matthias MüllerTeam:Jahr: 2021Förderung: This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 948679).Laufzeit: 2021 - 2025
-
Convex online optimization for dynamic systems controlLeitung: Prof. Dr.-Ing. Matthias MüllerTeam:Jahr: 2023Förderung: Deutsche Forschungsgemeinschaft (DFG) - 505182457Laufzeit: 2023 - 2025
-
ALeSCo Project 1: Neural network training via persistence of excitationLeitung: Prof. Dr.-Ing. Matthias Müller, Dr. Victor LopezJahr: 2025Förderung: Deutsche Forschungsgemeinschaft (DFG) - 535860958Laufzeit: 2025 - 2029
-
ALeSCo Research Unit: Active Learning for Systems and Control - Data Informativity, Uncertainty, and GuaranteesLeitung: Spokesman: Matthias MüllerTeam:Jahr: 2025Förderung: Deutsche Forschungsgemeinschaft (DFG) - 535860958Laufzeit: 2025 - 2029