Neurally Optimized Model Predictive Control for Autonomous Landing of a UAV on a moving platform - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2021

Neurally Optimized Model Predictive Control for Autonomous Landing of a UAV on a moving platform

Résumé

The uses of UAVs are developing rapidly and are now also used to extend human capabilities. Their control poses several problems that are still open because the system is subject to strong constraints. In this context landing on mobile platform is one of challenging one. In this paper an automated solution for autonomous landing of a UAV on a moving target based on camera vision only is proposed. The method uses a discrete Model Predictive Control (MPC) approach based on Linear Time Invariant (LTI) model. Constraints on inputs, input rates and states are considered and solved using a Neural Optimizer. Software In the Loop (SIL) experiments were performed and proved that the method is valid. Finally a tool-chain for implementation and testing using open-source components only is explained.
Fichier non déposé

Dates et versions

hal-03628918 , version 1 (03-04-2022)

Identifiants

Citer

Wojciech Strozecki, Saïd Mammar. Neurally Optimized Model Predictive Control for Autonomous Landing of a UAV on a moving platform. 18th IEEE International Conference on Networking, Sensing and Control (ICNSC 2021), Dec 2021, Xiamen, China. pp.1-6, ⟨10.1109/ICNSC52481.2021.9702230⟩. ⟨hal-03628918⟩
87 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More