MARL Control of On-Orbit Peg-in-Hole Assembly in Macro-Micro Robots

Hao Li; Xuchao Huang; Yibo Zhang; Quan An; Yao Zhang

doi:10.1016/j.ifacol.2025.11.471

MARL Control of On-Orbit Peg-in-Hole Assembly in Macro-Micro Robots

Hao Li^*
, Xuchao Huang^*
, Yibo Zhang
, Quan An^*
, Yao Zhang^*

^*Corresponding author for this work

School of Aerospace Engineering

Research output: Contribution to journal › Conference article › peer-review

Abstract

In recent years, on-orbit operations have become a research hotspot. Traditional on-orbit assembly using only robotic arms suffers from low precision. Therefore, this paper proposes a Macro-Micro Coordinated Robotic System (MMCRS) consisting of a robotic arm and a Stewart platform. The paper analyzes the kinematic characteristics of the MMCRS for the on-orbit peg-in-hole assembly problem and introduces a Multi-Agent Reinforcement Learning algorithm, Full-State MADDPG. A staged, variable-parameter reward function is designed, and the effectiveness of the algorithm is verified through simulations.

Original language	English
Pages (from-to)	2113-2118
Number of pages	6
Journal	IFAC-PapersOnLine
Volume	59
Issue number	20
DOIs	https://doi.org/10.1016/j.ifacol.2025.11.471
Publication status	Published - 1 Aug 2025
Event	23th IFAC Symposium on Automatic Control in Aerospace, ACA 2025 - Harbin, China Duration: 2 Aug 2025 → 6 Aug 2025

Keywords

Dynamic Reward Function
Macro-Micro Coordinated Robots
MARL
On-Orbit Assembly
Peg-in-Hole

Access to Document

10.1016/j.ifacol.2025.11.471

Cite this

@article{9052183c92744e6f822263f3328f78b3,

title = "MARL Control of On-Orbit Peg-in-Hole Assembly in Macro-Micro Robots",

abstract = "In recent years, on-orbit operations have become a research hotspot. Traditional on-orbit assembly using only robotic arms suffers from low precision. Therefore, this paper proposes a Macro-Micro Coordinated Robotic System (MMCRS) consisting of a robotic arm and a Stewart platform. The paper analyzes the kinematic characteristics of the MMCRS for the on-orbit peg-in-hole assembly problem and introduces a Multi-Agent Reinforcement Learning algorithm, Full-State MADDPG. A staged, variable-parameter reward function is designed, and the effectiveness of the algorithm is verified through simulations.",

keywords = "Dynamic Reward Function, Macro-Micro Coordinated Robots, MARL, On-Orbit Assembly, Peg-in-Hole",

author = "Hao Li and Xuchao Huang and Yibo Zhang and Quan An and Yao Zhang",

note = "Publisher Copyright: Copyright {\textcopyright} 2025 The Authors.; 23th IFAC Symposium on Automatic Control in Aerospace, ACA 2025 ; Conference date: 02-08-2025 Through 06-08-2025",

year = "2025",

month = aug,

day = "1",

doi = "10.1016/j.ifacol.2025.11.471",

language = "English",

volume = "59",

pages = "2113--2118",

journal = "IFAC-PapersOnLine",

issn = "2405-8963",

publisher = "Elsevier B.V.",

number = "20",

}

TY - JOUR

T1 - MARL Control of On-Orbit Peg-in-Hole Assembly in Macro-Micro Robots

AU - Li, Hao

AU - Huang, Xuchao

AU - Zhang, Yibo

AU - An, Quan

AU - Zhang, Yao

PY - 2025/8/1

Y1 - 2025/8/1

N2 - In recent years, on-orbit operations have become a research hotspot. Traditional on-orbit assembly using only robotic arms suffers from low precision. Therefore, this paper proposes a Macro-Micro Coordinated Robotic System (MMCRS) consisting of a robotic arm and a Stewart platform. The paper analyzes the kinematic characteristics of the MMCRS for the on-orbit peg-in-hole assembly problem and introduces a Multi-Agent Reinforcement Learning algorithm, Full-State MADDPG. A staged, variable-parameter reward function is designed, and the effectiveness of the algorithm is verified through simulations.

AB - In recent years, on-orbit operations have become a research hotspot. Traditional on-orbit assembly using only robotic arms suffers from low precision. Therefore, this paper proposes a Macro-Micro Coordinated Robotic System (MMCRS) consisting of a robotic arm and a Stewart platform. The paper analyzes the kinematic characteristics of the MMCRS for the on-orbit peg-in-hole assembly problem and introduces a Multi-Agent Reinforcement Learning algorithm, Full-State MADDPG. A staged, variable-parameter reward function is designed, and the effectiveness of the algorithm is verified through simulations.

KW - Dynamic Reward Function

KW - Macro-Micro Coordinated Robots

KW - MARL

KW - On-Orbit Assembly

KW - Peg-in-Hole

UR - https://www.scopus.com/pages/publications/105025898690

U2 - 10.1016/j.ifacol.2025.11.471

DO - 10.1016/j.ifacol.2025.11.471

M3 - Conference article

AN - SCOPUS:105025898690

SN - 2405-8963

VL - 59

SP - 2113

EP - 2118

JO - IFAC-PapersOnLine

JF - IFAC-PapersOnLine

IS - 20

T2 - 23th IFAC Symposium on Automatic Control in Aerospace, ACA 2025

Y2 - 2 August 2025 through 6 August 2025

ER -

MARL Control of On-Orbit Peg-in-Hole Assembly in Macro-Micro Robots

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this