Dependable Control and Decision group

DECODE – Dependable Control and Decision group

The DECODE (Dependable Control and Decision) group focuses on complex control systems, where coupling between subsystems, local sources of uncertainty, and structural changes can undermine reliability and generate dangerous behaviors. By leveraging tools from control theory, optimization, and machine learning, we contribute to the development of next-generation systems that are dependable, more autonomous and self adapt to changes in the internal dynamics and the surrounding environment.

The DECODE group is part of the Automatic Control Laboratory and the NCCR automation.

AA

Research topics

News

© 2022 EPFL

New PostDoc position in DECODE

— A PostDoc position is available within the Dependable Control and Decision group (DECODE) of EPFL (Switzerland) in the broad area of data-based control and learning. Candidates with a strong methodological background and motivated by research on theory and algorithms are encouraged to apply. More details can be found under Open Positions in the menu.

© 2022 EPFL

Contractive neural networks are robust against noise

— Robust Classification using Contractive Hamiltonian Neural ODEs

© 2022 EPFL

Shreyaa Raghavan joins the DECODE group

© 2022 EPFL

Contractivity promotes robustness of neural networks

— On Robust Classification using Contractive Hamiltonian Neural ODEs

© 2022 EPFL

New method for robust online joint state/input/parameter estimation

— Research on robust identification of electrical grid parameters lead us to devlop the AIRLS algorithm. It is an online method capable of tracking parameter changes in linear systems, even in the presence of noise and outliers in the measurements of all variables.

© 2022 EPFL

Port-Hamiltonian NNs for distributed control: oral presentation L4DC

— Our paper on port-Hamiltonian neural networks for distributed control with stability guarantees has been accepted for an oral presentation in the 4th Annual Learning for Dynamics & Control (L4DC) Conference. It will be held at Stanford University, USA, on June 23 – 24, 2022.

© 2022 EPFL

Master project in collaboration with Johnson Electric

— We are looking for motivated students for the master project " Actuator subjective noise detection via AI assisted time series analysis", to be developed in collaboration with Johnson Electric

© 2021 EPFL

Presentation on Bayesian distribution grid identification

— Watch the talk at CDC 2021 the 15th of December at 15:45, during the invited session on Machine Learning for Control of Power Systems.

© 2021 EPFL

New Award of the IEEE CSS Swiss Chapter

— We are happy to announce the first edition of the "IEEE Control Systems Society – Swiss Chapter Young Author Best Journal Paper Award". The award, sponsored by the Swiss Chapter of the IEEE Control System Society, addresses young researchers who have recently published papers in a journal of the society. More details about eligibility conditions are provided in this page.

All news

AA

Funding acknowledgment

Contact

Head of the group: Giancarlo Ferrari Trecate
Office: ME C2 398
Mail: [email protected]

Administration : Nicole Bouendin
Office : ME C2 389
Phone : +41 21 693 38 45
Fax : +41 21 693 2574
Mail : [email protected]


Dependable Control and Decision group (DECODE)
Laboratoire d’Automatique
Ecole Polytechnique Fédérale de Lausanne
EPFL STI IGM SCI-STI-GFT 
ME C2 398 (Bâtiment ME) , Station 9 
CH-1015 Lausanne
Switzerland