Reinforcement Learning based Health Aware Control Design - Application to Autonomous Systems

Satya Vinay Chavan MARTHI

Didier THEILLIOL , Mayank Shekhar JHA , Jean-Christophe PONSART

This is an exciting opportunity to participate in an interdisciplinary subject, combining autonomous vehicle behavior development and real-world experimentation, as part of a dynamic research team working on cutting-edge topics such as learning control, aerial and ground autonomous vehicles and comparative framework. This thesis, for which funding is guaranteed, is part of the ANR Self-Organizing, Smart and safe heterogeneous robots fleet by collective emergence SOS project (TSIA ANR call). SOS project brings together three research teams from two laboratories (CRIStAL and CRAN) and one SME (Lynxdrone) with the aim of proposing, designing and developing a mechanism for intelligent management of heterogeneous aerial and ground autonomous fleet of vehicles by collective emergence.

This thesis is part of the ANR SOS project, which focuses on collective behavior in aerial and ground autonomous vehicles. The aim is to explore within the learning paradigm the development of reconfiguration schemes for autonomous mobile control that consider the health of the system as well as predictions of future failures, thus guaranteeing the completion of the mission with a certain level of performance in terms of stability and safety. The thesis will focus on innovative approaches based in the framework of reinforcement learning for the design of safe control laws. This work will be based on simulation work on the one hand, and the production of a demonstrator on the other. The application framework is fire-fighting support.

Health Aware Control Design, reinforcement learning, autonomous mobile system, simulation

Control Identification Diagnosis