Deep Reinforcement Learning for the Visual Servoing Control of UAVs with FOV Constraint
Visual servoing is a control method that utilizes image feedback to control robot motion, and it has been widely applied in unmanned aerial vehicle (UAV) motion control. However, due to field-of-view (FOV) constraints, visual servoing still faces challenges, such as easy target loss and low control...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-06-01
|
| Series: | Drones |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2504-446X/7/6/375 |
| _version_ | 1797595277420396544 |
|---|---|
| author | Gui Fu Hongyu Chu Liwen Liu Linyi Fang Xinyu Zhu |
| author_facet | Gui Fu Hongyu Chu Liwen Liu Linyi Fang Xinyu Zhu |
| author_sort | Gui Fu |
| collection | DOAJ |
| description | Visual servoing is a control method that utilizes image feedback to control robot motion, and it has been widely applied in unmanned aerial vehicle (UAV) motion control. However, due to field-of-view (FOV) constraints, visual servoing still faces challenges, such as easy target loss and low control efficiency. To address these issues, visual servoing control for UAVs based on the deep reinforcement learning (DRL) method is proposed, which dynamically adjusts the servo gain in real time to avoid target loss and improve control efficiency. Firstly, a Markov model of visual servoing control for a UAV under field-of-view constraints is established, which consists ofquintuplet and considers the improvement of the control efficiency. Secondly, an improved deep Q-network (DQN) algorithm with a target network and experience replay is designed to solve the Markov model. In addition, two independent agents are designed to adjust the linear and angular velocity servo gains in order to enhance the control performance, respectively. In the simulation environment, the effectiveness of the proposed method was verified using a monocular camera. |
| first_indexed | 2024-03-11T02:33:57Z |
| format | Article |
| id | doaj.art-ac8e670c9911418a844a7b94009ad244 |
| institution | Directory Open Access Journal |
| issn | 2504-446X |
| language | English |
| last_indexed | 2024-03-11T02:33:57Z |
| publishDate | 2023-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Drones |
| spelling | doaj.art-ac8e670c9911418a844a7b94009ad2442023-11-18T10:04:12ZengMDPI AGDrones2504-446X2023-06-017637510.3390/drones7060375Deep Reinforcement Learning for the Visual Servoing Control of UAVs with FOV ConstraintGui Fu0Hongyu Chu1Liwen Liu2Linyi Fang3Xinyu Zhu4School of Information Engineering, Southwest University of Science and Technology, Mianyang 621010, ChinaSchool of Information Engineering, Southwest University of Science and Technology, Mianyang 621010, ChinaInstitute of Electronic and Electrical Engineering, Civil Aviation Flight University of China, Guanghan 618307, ChinaInstitute of Electronic and Electrical Engineering, Civil Aviation Flight University of China, Guanghan 618307, ChinaInstitute of Electronic and Electrical Engineering, Civil Aviation Flight University of China, Guanghan 618307, ChinaVisual servoing is a control method that utilizes image feedback to control robot motion, and it has been widely applied in unmanned aerial vehicle (UAV) motion control. However, due to field-of-view (FOV) constraints, visual servoing still faces challenges, such as easy target loss and low control efficiency. To address these issues, visual servoing control for UAVs based on the deep reinforcement learning (DRL) method is proposed, which dynamically adjusts the servo gain in real time to avoid target loss and improve control efficiency. Firstly, a Markov model of visual servoing control for a UAV under field-of-view constraints is established, which consists ofquintuplet and considers the improvement of the control efficiency. Secondly, an improved deep Q-network (DQN) algorithm with a target network and experience replay is designed to solve the Markov model. In addition, two independent agents are designed to adjust the linear and angular velocity servo gains in order to enhance the control performance, respectively. In the simulation environment, the effectiveness of the proposed method was verified using a monocular camera.https://www.mdpi.com/2504-446X/7/6/375field-of-view (FOV) constraintvisual servoingdeep reinforcement learningUAV |
| spellingShingle | Gui Fu Hongyu Chu Liwen Liu Linyi Fang Xinyu Zhu Deep Reinforcement Learning for the Visual Servoing Control of UAVs with FOV Constraint Drones field-of-view (FOV) constraint visual servoing deep reinforcement learning UAV |
| title | Deep Reinforcement Learning for the Visual Servoing Control of UAVs with FOV Constraint |
| title_full | Deep Reinforcement Learning for the Visual Servoing Control of UAVs with FOV Constraint |
| title_fullStr | Deep Reinforcement Learning for the Visual Servoing Control of UAVs with FOV Constraint |
| title_full_unstemmed | Deep Reinforcement Learning for the Visual Servoing Control of UAVs with FOV Constraint |
| title_short | Deep Reinforcement Learning for the Visual Servoing Control of UAVs with FOV Constraint |
| title_sort | deep reinforcement learning for the visual servoing control of uavs with fov constraint |
| topic | field-of-view (FOV) constraint visual servoing deep reinforcement learning UAV |
| url | https://www.mdpi.com/2504-446X/7/6/375 |
| work_keys_str_mv | AT guifu deepreinforcementlearningforthevisualservoingcontrolofuavswithfovconstraint AT hongyuchu deepreinforcementlearningforthevisualservoingcontrolofuavswithfovconstraint AT liwenliu deepreinforcementlearningforthevisualservoingcontrolofuavswithfovconstraint AT linyifang deepreinforcementlearningforthevisualservoingcontrolofuavswithfovconstraint AT xinyuzhu deepreinforcementlearningforthevisualservoingcontrolofuavswithfovconstraint |