File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: A two-stage reinforcement learning approach for multi-uav collision avoidance under imperfect sensing

TitleA two-stage reinforcement learning approach for multi-uav collision avoidance under imperfect sensing
Authors
KeywordsCollision avoidance
Robot sensing systems
Learning (artificial intelligence)
Training
Navigation
Issue Date2020
PublisherInstitute of Electrical and Electronics Engineers. The Journal's web site is located at https://www.ieee.org/membership-catalog/productdetail/showProductDetailPage.html?product=PER481-ELE
Citation
IEEE Robotics and Automation Letters, 2020, v. 5 n. 2, p. 3098-3105 How to Cite?
AbstractUnlike autonomous ground vehicles (AGVs), unmanned aerial vehicles (UAVs) have a higher dimensional configuration space, which makes the motion planning of multi-UAVs a challenging task. In addition, uncertainties and noises are more significant in UAV scenarios, which increases the difficulty of autonomous navigation for multi-UAV. In this letter, we proposed a two-stage reinforcement learning (RL) based multi-UAV collision avoidance approach without explicitly modeling the uncertainty and noise in the environment. Our goal is to train a policy to plan a collision-free trajectory by leveraging local noisy observations. However, the reinforcement learned collision avoidance policies usually suffer from high variance and low reproducibility, because unlike supervised learning, RL does not have a fixed training set with ground-truth labels. To address these issues, we introduced a two-stage training method for RL based collision avoidance. For the first stage, we optimize the policy using a supervised training method with a loss function that encourages the agent to follow the well-known reciprocal collision avoidance strategy. For the second stage, we use policy gradient to refine the policy. We validate our policy in a variety of simulated scenarios, and the extensive numerical simulations demonstrate that our policy can generate time-efficient and collision-free paths under imperfect sensing, and can well handle noisy local observations with unknown noise levels.
Persistent Identifierhttp://hdl.handle.net/10722/285105
ISSN
2023 Impact Factor: 4.6
2023 SCImago Journal Rankings: 2.119
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorWANG, D-
dc.contributor.authorFAN, T-
dc.contributor.authorHan, T-
dc.contributor.authorPan, J-
dc.date.accessioned2020-08-07T09:06:49Z-
dc.date.available2020-08-07T09:06:49Z-
dc.date.issued2020-
dc.identifier.citationIEEE Robotics and Automation Letters, 2020, v. 5 n. 2, p. 3098-3105-
dc.identifier.issn2377-3766-
dc.identifier.urihttp://hdl.handle.net/10722/285105-
dc.description.abstractUnlike autonomous ground vehicles (AGVs), unmanned aerial vehicles (UAVs) have a higher dimensional configuration space, which makes the motion planning of multi-UAVs a challenging task. In addition, uncertainties and noises are more significant in UAV scenarios, which increases the difficulty of autonomous navigation for multi-UAV. In this letter, we proposed a two-stage reinforcement learning (RL) based multi-UAV collision avoidance approach without explicitly modeling the uncertainty and noise in the environment. Our goal is to train a policy to plan a collision-free trajectory by leveraging local noisy observations. However, the reinforcement learned collision avoidance policies usually suffer from high variance and low reproducibility, because unlike supervised learning, RL does not have a fixed training set with ground-truth labels. To address these issues, we introduced a two-stage training method for RL based collision avoidance. For the first stage, we optimize the policy using a supervised training method with a loss function that encourages the agent to follow the well-known reciprocal collision avoidance strategy. For the second stage, we use policy gradient to refine the policy. We validate our policy in a variety of simulated scenarios, and the extensive numerical simulations demonstrate that our policy can generate time-efficient and collision-free paths under imperfect sensing, and can well handle noisy local observations with unknown noise levels.-
dc.languageeng-
dc.publisherInstitute of Electrical and Electronics Engineers. The Journal's web site is located at https://www.ieee.org/membership-catalog/productdetail/showProductDetailPage.html?product=PER481-ELE-
dc.relation.ispartofIEEE Robotics and Automation Letters-
dc.rightsIEEE Robotics and Automation Letters. Copyright © Institute of Electrical and Electronics Engineers.-
dc.rights©20xx IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.-
dc.subjectCollision avoidance-
dc.subjectRobot sensing systems-
dc.subjectLearning (artificial intelligence)-
dc.subjectTraining-
dc.subjectNavigation-
dc.titleA two-stage reinforcement learning approach for multi-uav collision avoidance under imperfect sensing-
dc.typeArticle-
dc.identifier.emailPan, J: jpan@cs.hku.hk-
dc.identifier.authorityPan, J=rp01984-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1109/LRA.2020.2974648-
dc.identifier.scopuseid_2-s2.0-85081588190-
dc.identifier.hkuros312130-
dc.identifier.volume5-
dc.identifier.issue2-
dc.identifier.spage3098-
dc.identifier.epage3105-
dc.identifier.isiWOS:000526521900003-
dc.publisher.placeUnited States-
dc.identifier.issnl2377-3766-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats