Details
| Original language | English |
|---|---|
| Pages (from-to) | 991-1006 |
| Number of pages | 16 |
| Journal | IEEE Transactions on Automatic Control |
| Volume | 68 |
| Issue number | 2 |
| Publication status | Published - 5 May 2022 |
Abstract
In this paper, we propose a novel learning-based model predictive control framework for nonlinear systems which is able to guarantee closed-loop learning of the controlled system. We consider a cost function that combines a general economic cost with a user-defined learning cost function that aims at incentivizing learning of the unknown system. In particular, due to the finite horizon of the MPC scheme and to the presence of disturbances, the open-loop trajectory usually differs from the closed-loop one. Such a mismatch causes existing learning-based MPC schemes to only show a learning phase in the open-loop prediction, without providing any formal guarantee on the actual closed-loop learning. In this paper, we show how existing MPC schemes can be easily modified in order to guarantee closed-loop learning of the system by including a suitable discount factor in the chosen learning cost function, and implementing an additional constraint in the original MPC scheme. We show that various techniques for online learning the system dynamics such as kinky inference methods, Gaussian processes, or parametric approaches, can be used within the proposed general framework.
Keywords
- Control systems, Cost function, Costs, Predictive control, Predictive models, Trajectory, Uncertainty, nonlinear systems, Constrained control, predictive control for nonlinear systems, closed loop learning
ASJC Scopus subject areas
- Engineering(all)
- Electrical and Electronic Engineering
- Engineering(all)
- Control and Systems Engineering
- Computer Science(all)
- Computer Science Applications
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: IEEE Transactions on Automatic Control, Vol. 68, No. 2, 05.05.2022, p. 991-1006.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - Guaranteed closed-loop learning in Model Predictive Control
AU - Soloperto, Raffaele
AU - Muller, Matthias A.
AU - Allgower, Frank
N1 - This work was supported by Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Grants AL 316/12-2 and MU 3929/1-2 - 279734922. The authors thank the International Max Planck Research School for Intelligent Systems (IMPRS-IS).
PY - 2022/5/5
Y1 - 2022/5/5
N2 - In this paper, we propose a novel learning-based model predictive control framework for nonlinear systems which is able to guarantee closed-loop learning of the controlled system. We consider a cost function that combines a general economic cost with a user-defined learning cost function that aims at incentivizing learning of the unknown system. In particular, due to the finite horizon of the MPC scheme and to the presence of disturbances, the open-loop trajectory usually differs from the closed-loop one. Such a mismatch causes existing learning-based MPC schemes to only show a learning phase in the open-loop prediction, without providing any formal guarantee on the actual closed-loop learning. In this paper, we show how existing MPC schemes can be easily modified in order to guarantee closed-loop learning of the system by including a suitable discount factor in the chosen learning cost function, and implementing an additional constraint in the original MPC scheme. We show that various techniques for online learning the system dynamics such as kinky inference methods, Gaussian processes, or parametric approaches, can be used within the proposed general framework.
AB - In this paper, we propose a novel learning-based model predictive control framework for nonlinear systems which is able to guarantee closed-loop learning of the controlled system. We consider a cost function that combines a general economic cost with a user-defined learning cost function that aims at incentivizing learning of the unknown system. In particular, due to the finite horizon of the MPC scheme and to the presence of disturbances, the open-loop trajectory usually differs from the closed-loop one. Such a mismatch causes existing learning-based MPC schemes to only show a learning phase in the open-loop prediction, without providing any formal guarantee on the actual closed-loop learning. In this paper, we show how existing MPC schemes can be easily modified in order to guarantee closed-loop learning of the system by including a suitable discount factor in the chosen learning cost function, and implementing an additional constraint in the original MPC scheme. We show that various techniques for online learning the system dynamics such as kinky inference methods, Gaussian processes, or parametric approaches, can be used within the proposed general framework.
KW - Control systems
KW - Cost function
KW - Costs
KW - Predictive control
KW - Predictive models
KW - Trajectory
KW - Uncertainty
KW - nonlinear systems
KW - Constrained control
KW - predictive control for nonlinear systems
KW - closed loop learning
UR - http://www.scopus.com/inward/record.url?scp=85130196043&partnerID=8YFLogxK
U2 - 10.1109/TAC.2022.3172453
DO - 10.1109/TAC.2022.3172453
M3 - Article
AN - SCOPUS:85130196043
VL - 68
SP - 991
EP - 1006
JO - IEEE Transactions on Automatic Control
JF - IEEE Transactions on Automatic Control
SN - 0018-9286
IS - 2
ER -