Skip to Main content Skip to Navigation
Theses

Intégration des préférences d'un opérateur dans les décisions d'un drone autonome et élicitation incrémentale de ces préférences

Abstract : A fully autonomous unmanned aerial vehicle (UAV) is an aircraft without a human pilot on board. It is consequently able to accomplish a mission without the intervention of a human operator and to make decisions in a totally autonomous way. This implies that the ground operator must have a high level of confidence in the decisions made by the UAV.The main objective of this thesis is therefore to propose a decision engine to be embedded in the autonomous UAV that guarantees a high level of operator confidence in the UAV's ability to make the "right" decisions. For this purpose, we propose a multi-level decision engine composed of two main decision levels. The first one monitors the state of the UAV and its environment to detect events that can disrupt the mission’s execution and trigger the second level. Once triggered, it allows to choose a highlevel action (landing, continuing,...) best adapted to the current situation from a set of possible actions. This engine also integrates the operator's preferences by using Multi-Criteria Decision Aiding models. They require a preliminary phase before the mission, where the operator's preferences are elicited, before being integrated into the UAV. To reduce the operator's effort during this phase, we propose an incremental elicitation process during which the questions submitted to the operator are deduced from the previous answers. This allows us to determine a model that accurately represents his or her preferences, while minimizing the number of questions.
Document type :
Theses
Complete list of metadatas

Cited literature [135 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-02536674
Contributor : Abes Star :  Contact
Submitted on : Wednesday, April 8, 2020 - 12:07:10 PM
Last modification on : Monday, October 19, 2020 - 10:55:42 AM

File

These-2019-MATHSTIC-Informatiq...
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-02536674, version 1

Citation

Arwa Khannoussi. Intégration des préférences d'un opérateur dans les décisions d'un drone autonome et élicitation incrémentale de ces préférences. Systèmes embarqués. Université de Bretagne occidentale - Brest, 2019. Français. ⟨NNT : 2019BRES0080⟩. ⟨tel-02536674⟩

Share

Metrics

Record views

201

Files downloads

58