Vision-Based Deep Reinforcement Learning of UAV-UGV Collaborative Landing Policy Using Automatic Curriculum
Collaborative autonomous landing of a quadrotor Unmanned Aerial Vehicle (UAV) on a moving Unmanned Ground Vehicle (UGV) presents challenges due to the need for accurate real-time tracking of the UGV and the adjustment for the landing policy. To address this challenge, we propose a progressive learni...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-11-01
|
| Series: | Drones |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2504-446X/7/11/676 |
| _version_ | 1797459510656237568 |
|---|---|
| author | Chang Wang Jiaqing Wang Changyun Wei Yi Zhu Dong Yin Jie Li |
| author_facet | Chang Wang Jiaqing Wang Changyun Wei Yi Zhu Dong Yin Jie Li |
| author_sort | Chang Wang |
| collection | DOAJ |
| description | Collaborative autonomous landing of a quadrotor Unmanned Aerial Vehicle (UAV) on a moving Unmanned Ground Vehicle (UGV) presents challenges due to the need for accurate real-time tracking of the UGV and the adjustment for the landing policy. To address this challenge, we propose a progressive learning framework for generating an optimal landing policy based on vision without the need of communication between the UAV and the UGV. First, we propose the Landing Vision System (LVS) to offer rapid localization and pose estimation of the UGV. Then, we design an Automatic Curriculum Learning (ACL) approach to learn the landing tasks under different conditions of UGV motions and wind interference. Specifically, we introduce a neural network-based difficulty discriminator to schedule the landing tasks according to their levels of difficulty. Our method achieves a higher landing success rate and accuracy compared with the state-of-the-art TD3 reinforcement learning algorithm. |
| first_indexed | 2024-03-09T16:53:24Z |
| format | Article |
| id | doaj.art-0421c960b49c43bcb324651413e7d547 |
| institution | Directory Open Access Journal |
| issn | 2504-446X |
| language | English |
| last_indexed | 2024-03-09T16:53:24Z |
| publishDate | 2023-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Drones |
| spelling | doaj.art-0421c960b49c43bcb324651413e7d5472023-11-24T14:38:12ZengMDPI AGDrones2504-446X2023-11-0171167610.3390/drones7110676Vision-Based Deep Reinforcement Learning of UAV-UGV Collaborative Landing Policy Using Automatic CurriculumChang Wang0Jiaqing Wang1Changyun Wei2Yi Zhu3Dong Yin4Jie Li5College of Intelligence Science and Technology, National University of Defense Technology, Changsha 410073, ChinaCollege of Mechanical and Electrical Engineering, Hohai University, Changzhou 213022, ChinaCollege of Mechanical and Electrical Engineering, Hohai University, Changzhou 213022, ChinaSchool of Computer Science, Nanjing Audit University, Nanjing 211800, ChinaCollege of Intelligence Science and Technology, National University of Defense Technology, Changsha 410073, ChinaCollege of Intelligence Science and Technology, National University of Defense Technology, Changsha 410073, ChinaCollaborative autonomous landing of a quadrotor Unmanned Aerial Vehicle (UAV) on a moving Unmanned Ground Vehicle (UGV) presents challenges due to the need for accurate real-time tracking of the UGV and the adjustment for the landing policy. To address this challenge, we propose a progressive learning framework for generating an optimal landing policy based on vision without the need of communication between the UAV and the UGV. First, we propose the Landing Vision System (LVS) to offer rapid localization and pose estimation of the UGV. Then, we design an Automatic Curriculum Learning (ACL) approach to learn the landing tasks under different conditions of UGV motions and wind interference. Specifically, we introduce a neural network-based difficulty discriminator to schedule the landing tasks according to their levels of difficulty. Our method achieves a higher landing success rate and accuracy compared with the state-of-the-art TD3 reinforcement learning algorithm.https://www.mdpi.com/2504-446X/7/11/676deep reinforcement learningautomatic curriculum learningUAV landing |
| spellingShingle | Chang Wang Jiaqing Wang Changyun Wei Yi Zhu Dong Yin Jie Li Vision-Based Deep Reinforcement Learning of UAV-UGV Collaborative Landing Policy Using Automatic Curriculum Drones deep reinforcement learning automatic curriculum learning UAV landing |
| title | Vision-Based Deep Reinforcement Learning of UAV-UGV Collaborative Landing Policy Using Automatic Curriculum |
| title_full | Vision-Based Deep Reinforcement Learning of UAV-UGV Collaborative Landing Policy Using Automatic Curriculum |
| title_fullStr | Vision-Based Deep Reinforcement Learning of UAV-UGV Collaborative Landing Policy Using Automatic Curriculum |
| title_full_unstemmed | Vision-Based Deep Reinforcement Learning of UAV-UGV Collaborative Landing Policy Using Automatic Curriculum |
| title_short | Vision-Based Deep Reinforcement Learning of UAV-UGV Collaborative Landing Policy Using Automatic Curriculum |
| title_sort | vision based deep reinforcement learning of uav ugv collaborative landing policy using automatic curriculum |
| topic | deep reinforcement learning automatic curriculum learning UAV landing |
| url | https://www.mdpi.com/2504-446X/7/11/676 |
| work_keys_str_mv | AT changwang visionbaseddeepreinforcementlearningofuavugvcollaborativelandingpolicyusingautomaticcurriculum AT jiaqingwang visionbaseddeepreinforcementlearningofuavugvcollaborativelandingpolicyusingautomaticcurriculum AT changyunwei visionbaseddeepreinforcementlearningofuavugvcollaborativelandingpolicyusingautomaticcurriculum AT yizhu visionbaseddeepreinforcementlearningofuavugvcollaborativelandingpolicyusingautomaticcurriculum AT dongyin visionbaseddeepreinforcementlearningofuavugvcollaborativelandingpolicyusingautomaticcurriculum AT jieli visionbaseddeepreinforcementlearningofuavugvcollaborativelandingpolicyusingautomaticcurriculum |