File Download
Supplementary
-
Citations:
- Appears in Collections:
Conference Paper: Flying through a narrow gap using neural network: an end-to-end planning and control approach
Title | Flying through a narrow gap using neural network: an end-to-end planning and control approach |
---|---|
Authors | |
Keywords | Deep Learning in Robotics and Automation Motion and Path Planning |
Issue Date | 2019 |
Publisher | IEEE/RSJ. |
Citation | IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS): Robots Connecting People, Macau, China, 4-8 November 2019 How to Cite? |
Abstract | In this paper, we investigate the problem of enabling a drone to fly through a tilted narrow gap, without a traditional planning and control pipeline. To this end, we propose an end-to-end policy network, which imitates from the traditional pipeline and is fine-tuned using reinforcement learning. Unlike previous works which plan dynamical feasible trajectories using motion primitives and track the generated trajectory by a geometric controller, our proposed method is an end-to-end approach which takes the flight scenario as input and directly outputs thrust-attitude control commands for the quadrotor. Key contributions of our paper are: 1) presenting an imitatereinforce training framework. 2) flying through a narrow gap using an end-to-end policy network, showing that learning based method can also address the highly dynamic control problem as the traditional pipeline does (see attached video1). 3) propose a robust imitation of an optimal trajectory generator using multilayer perceptrons. 4) show how reinforcement learning can improve the performance of imitation learning, and the potential to achieve higher performance over the model-based method. |
Description | WeBT3 Regular session, L1-R3: Learning for Motion and Path Planning II - Paper WeBT3.2 |
Persistent Identifier | http://hdl.handle.net/10722/274125 |
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Lin, J | - |
dc.contributor.author | Wang, L | - |
dc.contributor.author | Gao, F | - |
dc.contributor.author | Shen, S | - |
dc.contributor.author | Zhang, F | - |
dc.date.accessioned | 2019-08-18T14:55:35Z | - |
dc.date.available | 2019-08-18T14:55:35Z | - |
dc.date.issued | 2019 | - |
dc.identifier.citation | IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS): Robots Connecting People, Macau, China, 4-8 November 2019 | - |
dc.identifier.uri | http://hdl.handle.net/10722/274125 | - |
dc.description | WeBT3 Regular session, L1-R3: Learning for Motion and Path Planning II - Paper WeBT3.2 | - |
dc.description.abstract | In this paper, we investigate the problem of enabling a drone to fly through a tilted narrow gap, without a traditional planning and control pipeline. To this end, we propose an end-to-end policy network, which imitates from the traditional pipeline and is fine-tuned using reinforcement learning. Unlike previous works which plan dynamical feasible trajectories using motion primitives and track the generated trajectory by a geometric controller, our proposed method is an end-to-end approach which takes the flight scenario as input and directly outputs thrust-attitude control commands for the quadrotor. Key contributions of our paper are: 1) presenting an imitatereinforce training framework. 2) flying through a narrow gap using an end-to-end policy network, showing that learning based method can also address the highly dynamic control problem as the traditional pipeline does (see attached video1). 3) propose a robust imitation of an optimal trajectory generator using multilayer perceptrons. 4) show how reinforcement learning can improve the performance of imitation learning, and the potential to achieve higher performance over the model-based method. | - |
dc.language | eng | - |
dc.publisher | IEEE/RSJ. | - |
dc.relation.ispartof | IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) | - |
dc.subject | Deep Learning in Robotics and Automation | - |
dc.subject | Motion and Path Planning | - |
dc.title | Flying through a narrow gap using neural network: an end-to-end planning and control approach | - |
dc.type | Conference_Paper | - |
dc.identifier.email | Zhang, F: fuzhang@hku.hk | - |
dc.identifier.authority | Zhang, F=rp02460 | - |
dc.description.nature | postprint | - |
dc.identifier.hkuros | 301104 | - |