WAGNN: A Weighted Aggregation Graph Neural Network for robot skill learning

doi:10.1016/j.robot.2020.103555

CORC > 自动化研究所 > 中国科学院自动化研究所 > 复杂系统管理与控制国家重点实验室 > 机器人应用与理论组

	WAGNN: A Weighted Aggregation Graph Neural Network for robot skill learning
	Zhang, Fengyi 2,3; Liu, Zhiyong 1,2,3; Xiong, Fangzhou 2,3; Su, Jianhua 2; Qiao, Hong 1,2,3
刊名	ROBOTICS AND AUTONOMOUS SYSTEMS
	2020-08-01
卷号	130 页码:9
关键词	Skill transfer learning Serial structures Robot skill learning Graph Neural Network
ISSN号	0921-8890
DOI	10.1016/j.robot.2020.103555
通讯作者	Liu, Zhiyong(zhiyong.liu@ia.ac.cn)
英文摘要	Robotic skill learning suffers from the diversity and complexity of robotic tasks in continuous domains, making the learning of transferable skills one of the most challenging issues in this area, especially for the case where robots differ in terms of structure. Aiming at making the policy easier to be generalized or transferred, the graph neural networks (GNN) was previously employed to incorporate explicitly the robot structure into the policy network. In this paper, with the help of graph neural networks, we further investigate the problem of efficient learning transferable policies for robots with serial structure, which commonly appears in various robot bodies, such as robotic arms and the leg of centipede. Based on a kinematics analysis on the serial robotic structure, the policy network is improved by proposing a weighted information aggregation strategy. It is experimentally shown on different robotics structures that in a few-shot policy learning setting, the new aggregation strategy significantly improves the performance not only on the learning speed, but also on the control accuracy. (C) 2020 Elsevier B.V. All rights reserved.
资助项目	National Key Research and Development Plan of China[2017YFB1300202] ; NSFC, China[U1613213] ; NSFC, China[61375005] ; NSFC, China[61503383] ; NSFC, China[61210009] ; Strategic Priority Research Program of Chinese Academy of Science[XDB32050100] ; Dongguan core technology research frontier project, China[2019622101001]
WOS研究方向	Automation & Control Systems ; Computer Science ; Robotics
语种	英语
出版者	ELSEVIER
WOS记录号	WOS:000538810400003
资助机构	National Key Research and Development Plan of China ; NSFC, China ; Strategic Priority Research Program of Chinese Academy of Science ; Dongguan core technology research frontier project, China
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/39818]
专题	自动化研究所_复杂系统管理与控制国家重点实验室_机器人应用与理论组
通讯作者	Liu, Zhiyong
作者单位	1.Chinese Acad Sci, CAS Ctr Excellence Brain Sci & Intelligence Techn, Shanghai 200031, Peoples R China 2.Chinese Acad Sci, Inst Automat, State Key Lab Management & Control Complex Syst, Beijing 100190, Peoples R China 3.Univ Chinese Acad Sci UCAS, Sch Artificial Intelligence, Beijing 100049, Peoples R China
推荐引用方式 GB/T 7714	Zhang, Fengyi,Liu, Zhiyong,Xiong, Fangzhou,et al. WAGNN: A Weighted Aggregation Graph Neural Network for robot skill learning[J]. ROBOTICS AND AUTONOMOUS SYSTEMS,2020,130:9.
APA	Zhang, Fengyi,Liu, Zhiyong,Xiong, Fangzhou,Su, Jianhua,&Qiao, Hong.(2020).WAGNN: A Weighted Aggregation Graph Neural Network for robot skill learning.ROBOTICS AND AUTONOMOUS SYSTEMS,130,9.
MLA	Zhang, Fengyi,et al."WAGNN: A Weighted Aggregation Graph Neural Network for robot skill learning".ROBOTICS AND AUTONOMOUS SYSTEMS 130(2020):9.