自动化学报2026，Vol.52Issue(5)：1101-1115,15.DOI:10.16383/j.aas.c250686
基于PIRL的空间机械臂仿生智能抓取方法
Bionic Intelligent Grasping Method of Space Manipulators Based on Progressive Imitation-reinforcement Learning
摘要
Abstract
To address the challenges faced by space manipulators in performing autonomous grasping tasks of float-ing targets in microgravity environments,specifically the difficulties in sample acquisition,weak generalization cap-ability,and poor adaptation to dynamic disturbances,a bionic intelligence-integrated progressive imitation-rein-forcement learning method is proposed.First,based on expert demonstration data of human arm-hand collaborat-ive operations collected through teleoperation,a multi-layer perceptron(MLP)initial grasping strategy model is constructed,and bionic grasp training is conducted through behavior cloning;Next,the initial model is embedded into the high-fidelity Genesis space operation simulation environment,and the proximal policy optimization for grasping in space algorithm is employed for online fine-tuning of the grasping strategy.By leveraging a stacked ac-tion space and a staged reward mechanism,the method achieves collaborative optimization between expert prior knowledge and autonomous environmental exploration,effectively addressing the distribution shift defect in imita-tion learning and the sample efficiency bottleneck in reinforcement learning.Experimental results indicate that the proposed method achieves a grasp success rate of 89.5%under random target pose disturbances,an improvement of 14.5%compared to MLP-based imitation learning,significantly enhancing the robustness and environmental adapt-ability of the strategy in complex spatial scenes with target pose deviations.This provides a new technical solution for autonomous grasping of floating targets by space manipulators in microgravity environments.关键词
空间机械臂/漂浮目标自主抓取/仿生智能/强化学习Key words
space manipulator/autonomous grasping of floating targets/bionic intelligence/reinforcement learning引用本文复制引用
李连鹏,郭航,李明洋,张海博,徐拴锋,张冬浩..基于PIRL的空间机械臂仿生智能抓取方法[J].自动化学报,2026,52(5):1101-1115,15.基金项目
国家自然科学基金(62406032),北京市自然科学基金(4242036),空间智能控制技术全国重点实验室基金(HTKJ2025KL502016,2025-JCJQ-LB-065)资助 Supported by National Natural Science Foundation of China(62406032),Beijing Natural Science Foundation(4242036),and Fund of National Key Laboratory of Space Intelligent Control(HTKJ2025KL502016,2025-JCJQ-LB-065) （62406032）
