| Title: |
Learning System Dynamics from Sensory Input under Optimal Control Principles |
| Authors: |
Bounou, Oumayma; Ponce, Jean; Carpentier, Justin |
| Contributors: |
Département d'informatique - ENS-PSL (DI-ENS); École normale supérieure - Paris (ENS-PSL); Université Paris Sciences et Lettres (PSL)-Université Paris Sciences et Lettres (PSL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS); Models of visual object recognition and scene understanding (WILLOW); Université Paris Sciences et Lettres (PSL)-Université Paris Sciences et Lettres (PSL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-École normale supérieure - Paris (ENS-PSL); Université Paris Sciences et Lettres (PSL)-Université Paris Sciences et Lettres (PSL)-Institut National de Recherche en Informatique et en Automatique (Inria)-Centre National de la Recherche Scientifique (CNRS)-Centre Inria de Paris; Institut National de Recherche en Informatique et en Automatique (Inria); Courant Institute of Mathematical Sciences New York (CIMS); New York University New York (NYU); NYU System (NYU)-NYU System (NYU); Center for Data Science NYU (CDS); This work was supported in part by the HPC resources from GENCI-IDRIS (Grand 2023-AD011011263R3), the European Union through the AG-IMUS project (GA no. 101070165), the Louis Vuitton/ENS chair on artificial intelligence, the French government under the management of Agence Nationale de la Recherche as part of the ”Investissements d’avenir” program (reference ANR19-P3IA0001, PRAIRIE 3IA Institute), the ANR JCJC project NIMBLE (ANR-22-CE33-0008), and a Global Distinguished Professorship at the Courant Institute of Mathematical Sciences and the Center for Data Science at New York University.; ANR-22-CE33-0008,NIMBLE,Apprentissage et contrôle de modèles sensori-moteurs en robotique(2022) |
| Source: |
CDC 2024 Conference on Decision and Control ; https://inria.hal.science/hal-04661974 ; CDC 2024 Conference on Decision and Control, Dec 2024, Milan, Italy |
| Publisher Information: |
CCSD |
| Publication Year: |
2024 |
| Subject Terms: |
[MATH.MATH-OC]Mathematics [math]/Optimization and Control [math.OC]; [INFO.INFO-CV]Computer Science [cs]/Computer Vision and Pattern Recognition [cs.CV]; [SPI.AUTO]Engineering Sciences [physics]/Automatic |
| Subject Geographic: |
Milan; Italy |
| Description: |
International audience ; Identifying the underlying dynamics of actuated physical systems from sensory input is of high interest in control, robotics, and engineering in general. In the context of control problems, existing approaches decouple the construction of the feature space where the dynamics identification process occurs from the target control tasks, potentially leading to amismatch between feature and state spaces: the systems may notbe controllable in feature space, and synthesized controls maynot be applicable in state space. Borrowing from the Koopman formalism, we propose instead to learn an embedding of both the states and controls into a feature space where the dynamics are linear, and include the target control task in the learning objective in the form of a differentiable and robust optimal control problem. We validate the proposed approach withsimulation experiments using systems with non-linear dynamics, demonstrating that the controls obtained in feature space can be used to drive the corresponding physical systems and that the learned model can serve for future state prediction. |
| Document Type: |
conference object |
| Language: |
English |
| Availability: |
https://inria.hal.science/hal-04661974; https://inria.hal.science/hal-04661974v3/document; https://inria.hal.science/hal-04661974v3/file/cdc2024_cr.pdf |
| Rights: |
http://creativecommons.org/licenses/by/ ; info:eu-repo/semantics/OpenAccess |
| Accession Number: |
edsbas.73155CCE |
| Database: |
BASE |