Séminaires d’Andrei Braitor et Saverio Bolognani

Date : 27/04/2023
Catégorie(s) : ,
Lieu : Salle du conseil (L2S)

13h40–14h00 — Salle du conseil (L2S)

Adaptive droop control design with overcurrent protection for onboard DC microgrids in hybrid electric aircrafts

Andrei Braitor (L2S)

Abstract. In this talk, an adaptive nonlinear droop-based control approach is proposed for converter-based self-contained electrical power systems (EPS) designed for electric aircraft applications to ensure tight voltage regulation and accurate load power distribution among parallel sources. By taking into account the accurate nonlinear dynamic models of the power converters, a mathematical proof of an upper bound is guaranteed for the input current of each converter separately, by means of Lyapunov methods and ultimate boundedness theory. In essence, the adopted nonlinear droop-based controller introduces a virtual voltage and a virtual resistance in series with the inductance and parasitic resistance of each DC/DC boost converter. To verify the proposed controller performance and the underlying developed approach, simulation results of the low-voltage bus dynamics have been carried out for an onboard aircraft DC microgrid.

14h00–15h00 — Salle du conseil (L2S)

First-order and zeroth-order optimization algorithms as model-free feedback controllers

Saverio Bolognani (ETH Zurich)

Abstract. Many control problems in the engineering domain can be abstracted as multi-input multi-output steady-state regulation problems, where the system needs to be steered to an efficient operating point that is compatible with the operating limits of the plant. We will see that controllers for this class of problems can be designed by tapping into the tools of nonlinear optimization, and in particular first-order and zeroth-order methods. Interconnecting an iterative algorithm with a physical system requires however some care in order to guarantee well-posedness of the resulting dynamical system and closed-loop stability. On the other hand, this design approach yields robust controllers that are (almost) model-free and successfully tackle this complex steady-state control problems even in the presence of substantial uncertainty on the plant and on the unknown disturbances.

Biography. Saverio Bolognani received the Ph.D degree in Information Engineering from the University of Padova, Italy, in 2011. In 2006–2007, he was a visiting graduate student at the University of California at San Diego. In 2013–2014, he was a Postdoctoral Associate at the Laboratory for Information and Decision Systems of the Massachusetts Institute of Technology in Cambridge (MA). He is currently a Senior Scientist at the Automatic Control Laboratory at ETH Zurich. His research interests include the application of networked control system theory to power systems, cyber-physical systems, the intersection of nonlinear optimization with feedback control design, multi-agent systems, and game theory.