CASIA OpenIR  > 学术期刊  > Machine Intelligence Research
Skill Learning for Robotic Insertion Based on One-shot Demonstration and Reinforcement Learning
Ying Li1,2; De Xu1,2
发表期刊International Journal of Automation and Computing
ISSN1476-8186
2021
卷号18期号:3页码:457-467
摘要In this paper, an efficient skill learning framework is proposed for robotic insertion, based on one-shot demonstration and reinforcement learning. First, the robot action is composed of two parts: expert action and refinement action. A force Jacobian matrix is calibrated with only one demonstration, based on which stable and safe expert action can be generated. The deep deterministic policy gradients (DDPG) method is employed to learn the refinement action, which aims to improve the assembly efficiency. Second, an epis-ode-step exploration strategy is developed, which uses the expert action as a benchmark and adjusts the exploration intensity dynamically. A safety-efficiency reward function is designed for the compliant insertion. Third, to improve the adaptability with different components, a skill saving and selection mechanism is proposed. Several typical components are used to train the skill models. And the trained models and force Jacobian matrices are saved in a skill pool. Given a new component, the most appropriate model is selected from the skill pool according to the force Jacobian matrix and directly used to accomplish insertion tasks. Fourth, a simulation environment is established under the guidance of the force Jacobian matrix, which avoids tedious training process on real robotic systems. Simulation and experiments are conducted to validate the effectiveness of the proposed methods.
关键词Force Jacobian matrix one-shot demonstration dynamic exploration strategy insertion skill learning reinforcement
DOI10.1007/s11633-021-1290-3
引用统计
被引频次:6[WOS]   [WOS记录]     [WOS相关记录]
文献类型期刊论文
条目标识符http://ir.ia.ac.cn/handle/173211/44294
专题学术期刊_Machine Intelligence Research
作者单位1.Research Center of Precision Sensing and Control, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China
2.School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing 100049, China
第一作者单位精密感知与控制研究中心
推荐引用方式
GB/T 7714
Ying Li,De Xu. Skill Learning for Robotic Insertion Based on One-shot Demonstration and Reinforcement Learning[J]. International Journal of Automation and Computing,2021,18(3):457-467.
APA Ying Li,&De Xu.(2021).Skill Learning for Robotic Insertion Based on One-shot Demonstration and Reinforcement Learning.International Journal of Automation and Computing,18(3),457-467.
MLA Ying Li,et al."Skill Learning for Robotic Insertion Based on One-shot Demonstration and Reinforcement Learning".International Journal of Automation and Computing 18.3(2021):457-467.
条目包含的文件 下载所有文件
文件名称/大小 文献类型 版本类型 开放类型 使用许可
IJAC-2020-10-274.pdf(1450KB)期刊论文出版稿开放获取CC BY-NC-SA浏览 下载
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[Ying Li]的文章
[De Xu]的文章
百度学术
百度学术中相似的文章
[Ying Li]的文章
[De Xu]的文章
必应学术
必应学术中相似的文章
[Ying Li]的文章
[De Xu]的文章
相关权益政策
暂无数据
收藏/分享
文件名: IJAC-2020-10-274.pdf
格式: Adobe PDF
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。