Learning-based robot assembly method for peg insertion tasks on inclined hole using time-series force information

作     者：Zhifei Shen Zhiyong Jiang Jingwang Zhang Jun Wu Qiuguo Zhu 

作者机构：Polytechnic InstituteZhejiang UniversityHangzhou 310015China Robotics Engineering Centerthe 21st Research Institute of China Electronics Technology Group CorporationShanghai 200233China Department of Control Science and EngineeringZhejiang UniversityHangzhou 310027China 

出 版 物：《Biomimetic Intelligence & Robotics》 (仿生智能与机器人(英文))

年 卷 期：2025年第5卷第1期

页      面：116-123页

核心收录：

学科分类：080202[工学-机械电子工程] 08[工学] 0804[工学-仪器科学与技术] 0802[工学-机械工程] 

基　　金：supported by the National Key R&D Program of China(2022YFB4701502) the“Leading Goose”R&D Program of Zhejiang(2023C01177) the 2035 Key Technological Innovation Program of Ningbo City(2024Z300) 

主　　题：Manipulator assembly Peg insertion Force information Contact state Dual-arm 

摘      要：This paper presents a novel method for learning force-aware robot assembly skills,specifically targeting the peg insertion task on inclined *** the peg insertion task involving inclined holes,we employ one-dimensional convolutional networks(1DCNN)and gated recurrent units(GRU)to extract features from the time-series force information during the assembly process,thereby identifying different contact states between the peg and the *** to the identification of contact states,corresponding pose adjustments are executed,and overall smooth interaction is ensured through admittance *** assembly process is dynamically adjusted using a state machine to fine-tune admittance control parameters and seamlessly switch the assembly *** the utilization of dual-arm clamping,we conduct key unlocking experiments on bases inclined at varying *** results demonstrate that the proposed method significantly improves the accuracy and success rate of state recognition compared to previous methods.