Skip to Main content Skip to Navigation
Theses

Bracketing largest invariant sets of dynamical systems : an application to drifting underwater robots in ocean currents

Abstract : The proof of safety of robotic systems is afundamental issue for the development of robotics. Itconsists, for instance, in verifying that a robot controllaw will always satisfy a set of constraints. More gen-erally, we will be interested here in the verificationof the properties of dynamical systems that allow tomodel the evolution of a robot.The main contribution of this thesis is to providea new way of bracketing invariant sets of dynami-cal systems. To this end, a new abstract domain, themazes, and new algorithms are presented. It is alsoshown, through many examples, how classic valida-tion problems can be translated into a problem ofbracketing invariant sets. Finally, the results are ex-tended to the bracket of viability kernels.This thesis is also based on an application in un-derwater robotics. The main idea is to use oceancurrents so that an underwater robot can efficientlytravel long distances. A new kind of low-cost au-tonomous robot has been developed for this type ofmission. This new hybrid profiling float is able to reg-ulate its depth with a new regulation law, but also tocorrect its trajectory using auxiliary thrusters. Theyallow the robot to choose the right flow of current tobe used. The previously introduced validation toolsare applied to validate the robot and the missionsafety. Experiments in real conditions also enabledthe prototype to be validated.
Document type :
Theses
Complete list of metadata

https://tel.archives-ouvertes.fr/tel-03174429
Contributor : ABES STAR :  Contact
Submitted on : Friday, March 19, 2021 - 10:57:13 AM
Last modification on : Monday, April 4, 2022 - 9:28:17 AM

File

2019_thomas_lemezo.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-03174429, version 1

Citation

Thomas Le Mézo. Bracketing largest invariant sets of dynamical systems : an application to drifting underwater robots in ocean currents. Robotics [cs.RO]. ENSTA Bretagne - École nationale supérieure de techniques avancées Bretagne, 2019. English. ⟨NNT : 2019ENTA0012⟩. ⟨tel-03174429⟩

Share

Metrics

Record views

148

Files downloads

31