Residual Reinforcement Learning for Motion Control of a Bionic Exploration Robot-RoboDact

doi:10.1109/TIM.2023.3282297

CORC > 自动化研究所 > 中国科学院自动化研究所 > 多模态人工智能系统全国重点实验室

	Residual Reinforcement Learning for Motion Control of a Bionic Exploration Robot-RoboDact
	Zhang, Tiandong 4,5; Wang, Rui 5; Wang, Shuo 1,2,4,5; Wang, Yu 5; Zheng, Gang 3; Tan, Min
刊名	IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
	2023
卷号	72 页码:13
关键词	Active disturbance rejection control (ADRC) bionic exploration robot motion control residual reinforcement learning (RRL) soft actor-critic (SAC)
ISSN号	0018-9456
DOI	10.1109/TIM.2023.3282297
通讯作者	Wang, Rui(rwang5212@ia.ac.cn)
英文摘要	This article aims to investigate the motion control method of a bionic underwater exploration robot (RoboDact). The robot is equipped with a double-joint tail fin and two undulating pectoral fins to obtain good mobility and stability. The hybrid propulsion mode helps perform stable and effective underwater exploration and measurement. To coordinate these two kinds of bionic propulsion fins and address the challenge of measurement noises and external disturbances during underwater exploration, a novel residual reinforcement learning method with parameter randomization (PR-RRL) is proposed. The control strategy is a weighted superposition of a feedback controller and a residual controller. The observation feedback controller based on active disturbance rejection control (ADRC) is adapted to improve stability and convergence. And the residual controller based on the soft actor-critic (SAC) algorithm is adapted to improve adaptability to uncertainties and disturbances. Moreover, the parameter randomization training strategy is proposed for adapting natural complicated scenarios by randomizing the partial dynamics of the underwater exploration robot during the training phase. Finally, the feasibility and efficacy of the presented motion control method are validated by comprehensive simulation tests and RoboDact prototype physical experiments.
资助项目	STI 2030-Major Projects[2021ZD0114504] ; National Natural Science Foundation of China[62276253] ; National Natural Science Foundation of China[62203435] ; National Natural Science Foundation of China[62122087] ; Beijing Natural Science Foundation[4222055] ; Beijing Natural Science Foundation[4222056] ; Youth Innovation Promotion Association CAS[2020137] ; Scientific Research Program of Beijing Municipal Commission of Education-Natural Science Foundation of Beijing[KZ202210017024]
WOS关键词	FISH ; IMPLEMENTATION ; MANEUVERS
WOS研究方向	Engineering ; Instruments & Instrumentation
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:001012772500001
资助机构	STI 2030-Major Projects ; National Natural Science Foundation of China ; Beijing Natural Science Foundation ; Youth Innovation Promotion Association CAS ; Scientific Research Program of Beijing Municipal Commission of Education-Natural Science Foundation of Beijing
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/53762]
专题	多模态人工智能系统全国重点实验室
通讯作者	Wang, Rui
作者单位	1.Chinese Acad Sci, Ctr Excellence Brain Sci & Intelligence Technol, F-59000 Lille, Peoples R China 2.Univ Chinese Acad Sci, Sch Artificial Intelligence, Shanghai 200031, Peoples R China 3.Univ Lille, CRIStAL Ctr Rech Informat Signal & Automat Lille, Cent Lille, Lille 100190, France 4.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China 5.Chinese Acad Sci, Inst Automat, State Key Lab Multimodal Artificial Intelligence, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Zhang, Tiandong,Wang, Rui,Wang, Shuo,et al. Residual Reinforcement Learning for Motion Control of a Bionic Exploration Robot-RoboDact[J]. IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT,2023,72:13.
APA	Zhang, Tiandong,Wang, Rui,Wang, Shuo,Wang, Yu,Zheng, Gang,&Tan, Min.(2023).Residual Reinforcement Learning for Motion Control of a Bionic Exploration Robot-RoboDact.IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT,72,13.
MLA	Zhang, Tiandong,et al."Residual Reinforcement Learning for Motion Control of a Bionic Exploration Robot-RoboDact".IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 72(2023):13.