Skip to Main content Skip to Navigation
New interface
Theses

Commande sous contraintes de systèmes dynamiques multi-agents

Abstract : The goal of this thesis is to propose solutions for the optimal control of multi-agent dynamical systems under constraints. Elements from control theory and optimization are merged together in order to provide useful tools which are further applied to different problems involving multi-agent formations. The thesis considers the challenging case of agents subject to dynamical constraints. To deal with these issues, well established concepts like set-theory, differential flatness, Model Predictive Control (MPC), Mixed-Integer Programming (MIP) are adapted and enhanced. Using these theoretical notions, the thesis concentrates on understanding the geometrical properties of the multi-agent group formation and on providing a novel synthesis framework which exploits the group structure. In particular, the formation design and the collision avoidance conditions are casted as geometrical problems and optimization-based procedures are developed to solve them. Moreover, considerable advances in this direction are obtained by efficiently using MIP techniques (in order to derive an efficient description of the non-convex, non-connected feasible region which results from multi-agent collision and obstacle avoidance constraints) and stability properties (in order to analyze the uniqueness and existence of formation configurations). Lastly, some of the obtained theoretical results are applied on a challenging practical application. A novel combination of MPC and differential flatness (for reference generation) is used for the flight control of Unmanned Aerial Vehicles (UAVs).
Document type :
Theses
Complete list of metadata

Cited literature [248 references]  Display  Hide  Download

https://theses.hal.science/tel-00783221
Contributor : ABES STAR :  Contact
Submitted on : Thursday, January 31, 2013 - 4:37:11 PM
Last modification on : Sunday, June 26, 2022 - 11:58:12 AM
Long-term archiving on: : Wednesday, May 1, 2013 - 3:57:51 AM

File

Prodan_Ionela_thesis_VF_20.12....
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-00783221, version 1

Citation

Ionela Prodan. Commande sous contraintes de systèmes dynamiques multi-agents. Autre. Supélec, 2012. Français. ⟨NNT : 2012SUPL0019⟩. ⟨tel-00783221⟩

Share

Metrics

Record views

661

Files downloads

1489