Collège Doctoral de l'Université de Grenoble

Programme :
Abstract :

This course is dedicated to the application ofoptimization-based control strategies for complex systems.Its content offers a geometric viewpoint for various control problems which appear inmotion planningfor autonomous (aerial, aquatic) vehiclesandinterconnected systemslikeDC microgridsorwaternetworks.Why optimization-based control?Today’s systems have complex architectures, are bounded byheterogeneous constraints and have to respect challenging operational costs. Thus, optimization-basedapproaches, which fully exploit changes in mission, initial condition and environment, are becomingubiquitous within the community.Why a geometric viewpoint?Most control problems can be better understood if seen from a geo-metrical viewpoint. Such an interpretation may lead to new theoretical insights and streamline theirrepresentation. Furthermore, there are many tools which can be readily applied such as set-basedmethods and combinatorial results.The tools the students will acquire during this course in optimization-based control (via distributedand hierarchical approaches) will be used for several application benchmarks such as control andcoordination of multiple (aerial, aquatic) drones and energy management in DC microgrid systems.

Main instructor:
Assoc. Prof. HDRIonela Prodan, Univ. Grenoble Alpes, Grenoble INP, LCIS (Laboratory ofConception and Integration of Systems), Valence, France (see alsoGoogle scholar).

Invited speakers (2021-2022):
Assoc. Prof.Francesca Boem, UCL (University College London), Dept. of Electronic and ElectricalEngineering, Faculty of Engineering Science, London, UK (see alsoGoogle Scholar).Prof.Florin Stoican, UPB (University Politehnica Bucharest), Dept. of Automatic Control andSystems Engineering, Bucharest, Romania (see alsoGoogle Scholar).

Audience/prerequisites : The course is intended for PhD students in applied science with a major in control. A basic knowledgein control theory (state space representation) and Matlab/Python/C programming are required.

Remarks/organization:
This is a 21 hours course which is divided into 12 hours of theoretical lectures and 9 hours oflaboratory sessions dedicated to Matlab/Python simulations. The applications range from multi-agent obstacle and collision avoidance, trajectory planning and tracking for autonomous vehicles tohierarchical control for a DC microgrid system.