No Cover Image

Journal article 22 views 4 downloads

GCTD3: Modeling of Bipedal Locomotion by Combination of TD3 Algorithms and Graph Convolutional Network

Khoi Phan Bui Orcid Logo, GIANG NGUYEN, Dat Nguyen Ngoc Orcid Logo

Applied Sciences, Volume: 12, Issue: 6, Start page: 2948

Swansea University Author: GIANG NGUYEN

  • 71636.VoR.pdf

    PDF | Version of Record

    © 2022 by the authors. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license.

    Download (5.04MB)

Check full text

DOI (Published version): 10.3390/app12062948

Abstract

In recent years, there has been a lot of research using reinforcement learning algorithms to train 2-legged robots to move, but there are still many challenges. The authors propose the GCTD3 method, which takes the idea of using Graph Convolutional Networks to represent the kinematic link features o...

Full description

Published in: Applied Sciences
ISSN: 2076-3417
Published: MDPI AG 2022
Online Access: Check full text

URI: https://cronfa.swan.ac.uk/Record/cronfa71636
first_indexed 2026-03-17T22:01:36Z
last_indexed 2026-04-15T04:47:33Z
id cronfa71636
recordtype SURis
fullrecord <?xml version="1.0"?><rfc1807><datestamp>2026-04-14T15:23:22.7504748</datestamp><bib-version>v2</bib-version><id>71636</id><entry>2026-03-17</entry><title>GCTD3: Modeling of Bipedal Locomotion by Combination of TD3 Algorithms and Graph Convolutional Network</title><swanseaauthors><author><sid>2c72c4aaae11f710d293ec9ae69ce4e9</sid><firstname>GIANG</firstname><surname>NGUYEN</surname><name>GIANG NGUYEN</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2026-03-17</date><abstract>In recent years, there has been a lot of research using reinforcement learning algorithms to train 2-legged robots to move, but there are still many challenges. The authors propose the GCTD3 method, which takes the idea of using Graph Convolutional Networks to represent the kinematic link features of the robot, and combines this with the Twin-Delayed Deep Deterministic Policy Gradient algorithm to train the robot to move. Graph Convolutional Networks are very effective in graph-structured problems such as the connection of the joints of the human-like robots. The GCTD3 method shows better results on the motion trajectories of the bipedal robot joints compared with other reinforcement learning algorithms such as Twin-Delayed Deep Deterministic Policy Gradient, Deep Deterministic Policy Gradient and Soft Actor Critic. This research is implemented on a 2-legged robot model with six independent joint coordinates through the Robot Operating System and Gazebo simulator.</abstract><type>Journal Article</type><journal>Applied Sciences</journal><volume>12</volume><journalNumber>6</journalNumber><paginationStart>2948</paginationStart><paginationEnd/><publisher>MDPI AG</publisher><placeOfPublication/><isbnPrint/><isbnElectronic/><issnPrint/><issnElectronic>2076-3417</issnElectronic><keywords>GCTD3; GCN; TD3; ROS; reward function; bipedal robot</keywords><publishedDay>14</publishedDay><publishedMonth>3</publishedMonth><publishedYear>2022</publishedYear><publishedDate>2022-03-14</publishedDate><doi>10.3390/app12062948</doi><url/><notes/><college>COLLEGE NANME</college><CollegeCode>COLLEGE CODE</CollegeCode><institution>Swansea University</institution><apcterm/><funders>N/A</funders><projectreference/><lastEdited>2026-04-14T15:23:22.7504748</lastEdited><Created>2026-03-17T18:59:59.0368867</Created><path><level id="1">Faculty of Science and Engineering</level><level id="2">School of Mathematics and Computer Science - Computer Science</level></path><authors><author><firstname>Khoi Phan</firstname><surname>Bui</surname><orcid>0000-0002-1287-8879</orcid><order>1</order></author><author><firstname>GIANG</firstname><surname>NGUYEN</surname><order>2</order></author><author><firstname>Dat Nguyen</firstname><surname>Ngoc</surname><orcid>0000-0002-5842-9606</orcid><order>3</order></author></authors><documents><document><filename>71636__36507__0e4fff730104441b96e8fcdf9b25ab73.pdf</filename><originalFilename>71636.VoR.pdf</originalFilename><uploaded>2026-04-14T15:20:56.3743162</uploaded><type>Output</type><contentLength>5289793</contentLength><contentType>application/pdf</contentType><version>Version of Record</version><cronfaStatus>true</cronfaStatus><documentNotes>&#xA9; 2022 by the authors. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license.</documentNotes><copyrightCorrect>true</copyrightCorrect><language>eng</language><licence>https://creativecommons.org/licenses/by/4.0/</licence></document></documents><OutputDurs/></rfc1807>
spelling 2026-04-14T15:23:22.7504748 v2 71636 2026-03-17 GCTD3: Modeling of Bipedal Locomotion by Combination of TD3 Algorithms and Graph Convolutional Network 2c72c4aaae11f710d293ec9ae69ce4e9 GIANG NGUYEN GIANG NGUYEN true false 2026-03-17 In recent years, there has been a lot of research using reinforcement learning algorithms to train 2-legged robots to move, but there are still many challenges. The authors propose the GCTD3 method, which takes the idea of using Graph Convolutional Networks to represent the kinematic link features of the robot, and combines this with the Twin-Delayed Deep Deterministic Policy Gradient algorithm to train the robot to move. Graph Convolutional Networks are very effective in graph-structured problems such as the connection of the joints of the human-like robots. The GCTD3 method shows better results on the motion trajectories of the bipedal robot joints compared with other reinforcement learning algorithms such as Twin-Delayed Deep Deterministic Policy Gradient, Deep Deterministic Policy Gradient and Soft Actor Critic. This research is implemented on a 2-legged robot model with six independent joint coordinates through the Robot Operating System and Gazebo simulator. Journal Article Applied Sciences 12 6 2948 MDPI AG 2076-3417 GCTD3; GCN; TD3; ROS; reward function; bipedal robot 14 3 2022 2022-03-14 10.3390/app12062948 COLLEGE NANME COLLEGE CODE Swansea University N/A 2026-04-14T15:23:22.7504748 2026-03-17T18:59:59.0368867 Faculty of Science and Engineering School of Mathematics and Computer Science - Computer Science Khoi Phan Bui 0000-0002-1287-8879 1 GIANG NGUYEN 2 Dat Nguyen Ngoc 0000-0002-5842-9606 3 71636__36507__0e4fff730104441b96e8fcdf9b25ab73.pdf 71636.VoR.pdf 2026-04-14T15:20:56.3743162 Output 5289793 application/pdf Version of Record true © 2022 by the authors. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license. true eng https://creativecommons.org/licenses/by/4.0/
title GCTD3: Modeling of Bipedal Locomotion by Combination of TD3 Algorithms and Graph Convolutional Network
spellingShingle GCTD3: Modeling of Bipedal Locomotion by Combination of TD3 Algorithms and Graph Convolutional Network
GIANG NGUYEN
title_short GCTD3: Modeling of Bipedal Locomotion by Combination of TD3 Algorithms and Graph Convolutional Network
title_full GCTD3: Modeling of Bipedal Locomotion by Combination of TD3 Algorithms and Graph Convolutional Network
title_fullStr GCTD3: Modeling of Bipedal Locomotion by Combination of TD3 Algorithms and Graph Convolutional Network
title_full_unstemmed GCTD3: Modeling of Bipedal Locomotion by Combination of TD3 Algorithms and Graph Convolutional Network
title_sort GCTD3: Modeling of Bipedal Locomotion by Combination of TD3 Algorithms and Graph Convolutional Network
author_id_str_mv 2c72c4aaae11f710d293ec9ae69ce4e9
author_id_fullname_str_mv 2c72c4aaae11f710d293ec9ae69ce4e9_***_GIANG NGUYEN
author GIANG NGUYEN
author2 Khoi Phan Bui
GIANG NGUYEN
Dat Nguyen Ngoc
format Journal article
container_title Applied Sciences
container_volume 12
container_issue 6
container_start_page 2948
publishDate 2022
institution Swansea University
issn 2076-3417
doi_str_mv 10.3390/app12062948
publisher MDPI AG
college_str Faculty of Science and Engineering
hierarchytype
hierarchy_top_id facultyofscienceandengineering
hierarchy_top_title Faculty of Science and Engineering
hierarchy_parent_id facultyofscienceandengineering
hierarchy_parent_title Faculty of Science and Engineering
department_str School of Mathematics and Computer Science - Computer Science{{{_:::_}}}Faculty of Science and Engineering{{{_:::_}}}School of Mathematics and Computer Science - Computer Science
document_store_str 1
active_str 0
description In recent years, there has been a lot of research using reinforcement learning algorithms to train 2-legged robots to move, but there are still many challenges. The authors propose the GCTD3 method, which takes the idea of using Graph Convolutional Networks to represent the kinematic link features of the robot, and combines this with the Twin-Delayed Deep Deterministic Policy Gradient algorithm to train the robot to move. Graph Convolutional Networks are very effective in graph-structured problems such as the connection of the joints of the human-like robots. The GCTD3 method shows better results on the motion trajectories of the bipedal robot joints compared with other reinforcement learning algorithms such as Twin-Delayed Deep Deterministic Policy Gradient, Deep Deterministic Policy Gradient and Soft Actor Critic. This research is implemented on a 2-legged robot model with six independent joint coordinates through the Robot Operating System and Gazebo simulator.
published_date 2022-03-14T07:39:57Z
_version_ 1862698792221933568
score 11.102318

Similar Items