Quadrotor Autonomous Navigation in Semi-Known Environments Based on Deep Reinforcement Learning

In the application scenarios of quadrotors, it is expected that only part of the obstacles can be identified and located in advance. In order to make quadrotors fly safely in this situation, we present a deep reinforcement learning-based framework to realize autonomous navigation in semi-known envir...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Jiajun Ou, Xiao Guo, Wenjie Lou, Ming Zhu
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	unmanned aerial vehicle path planning obstacle avoidance deep reinforcement learning Science Q
Acceso en línea:	https://doaj.org/article/b9c4b23f47684bb7b1daa3caf6ff2575
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:b9c4b23f47684bb7b1daa3caf6ff2575
record_format	dspace
spelling	oai:doaj.org-article:b9c4b23f47684bb7b1daa3caf6ff25752021-11-11T18:53:58ZQuadrotor Autonomous Navigation in Semi-Known Environments Based on Deep Reinforcement Learning10.3390/rs132143302072-4292https://doaj.org/article/b9c4b23f47684bb7b1daa3caf6ff25752021-10-01T00:00:00Zhttps://www.mdpi.com/2072-4292/13/21/4330https://doaj.org/toc/2072-4292In the application scenarios of quadrotors, it is expected that only part of the obstacles can be identified and located in advance. In order to make quadrotors fly safely in this situation, we present a deep reinforcement learning-based framework to realize autonomous navigation in semi-known environments. Specifically, the proposed framework utilizes the dueling double deep recurrent Q-learning, which can implement global path planning with the obstacle map as input. Moreover, the proposed framework combined with contrastive learning-based feature extraction can conduct real-time autonomous obstacle avoidance with monocular vision effectively. The experimental results demonstrate that our framework exhibits remarkable performance for both global path planning and autonomous obstacle avoidance.Jiajun OuXiao GuoWenjie LouMing ZhuMDPI AGarticleunmanned aerial vehiclepath planningobstacle avoidancedeep reinforcement learningScienceQENRemote Sensing, Vol 13, Iss 4330, p 4330 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	unmanned aerial vehicle path planning obstacle avoidance deep reinforcement learning Science Q
spellingShingle	unmanned aerial vehicle path planning obstacle avoidance deep reinforcement learning Science Q Jiajun Ou Xiao Guo Wenjie Lou Ming Zhu Quadrotor Autonomous Navigation in Semi-Known Environments Based on Deep Reinforcement Learning
description	In the application scenarios of quadrotors, it is expected that only part of the obstacles can be identified and located in advance. In order to make quadrotors fly safely in this situation, we present a deep reinforcement learning-based framework to realize autonomous navigation in semi-known environments. Specifically, the proposed framework utilizes the dueling double deep recurrent Q-learning, which can implement global path planning with the obstacle map as input. Moreover, the proposed framework combined with contrastive learning-based feature extraction can conduct real-time autonomous obstacle avoidance with monocular vision effectively. The experimental results demonstrate that our framework exhibits remarkable performance for both global path planning and autonomous obstacle avoidance.
format	article
author	Jiajun Ou Xiao Guo Wenjie Lou Ming Zhu
author_facet	Jiajun Ou Xiao Guo Wenjie Lou Ming Zhu
author_sort	Jiajun Ou
title	Quadrotor Autonomous Navigation in Semi-Known Environments Based on Deep Reinforcement Learning
title_short	Quadrotor Autonomous Navigation in Semi-Known Environments Based on Deep Reinforcement Learning
title_full	Quadrotor Autonomous Navigation in Semi-Known Environments Based on Deep Reinforcement Learning
title_fullStr	Quadrotor Autonomous Navigation in Semi-Known Environments Based on Deep Reinforcement Learning
title_full_unstemmed	Quadrotor Autonomous Navigation in Semi-Known Environments Based on Deep Reinforcement Learning
title_sort	quadrotor autonomous navigation in semi-known environments based on deep reinforcement learning
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/b9c4b23f47684bb7b1daa3caf6ff2575
work_keys_str_mv	AT jiajunou quadrotorautonomousnavigationinsemiknownenvironmentsbasedondeepreinforcementlearning AT xiaoguo quadrotorautonomousnavigationinsemiknownenvironmentsbasedondeepreinforcementlearning AT wenjielou quadrotorautonomousnavigationinsemiknownenvironmentsbasedondeepreinforcementlearning AT mingzhu quadrotorautonomousnavigationinsemiknownenvironmentsbasedondeepreinforcementlearning
_version_	1718431667763281920

Quadrotor Autonomous Navigation in Semi-Known Environments Based on Deep Reinforcement Learning

Ejemplares similares