Learning and Generalizing Cooperative Manipulation Skills Using Parametric Dynamic Movement Primitives

作     者：Kim, Hyoin Oh, Changsuk Jang, Inkyu Park, Sungyong Seo, Hoseong Kim, H. Jin 

作者机构：Samsung Elect Mechatron Res & Dev Ctr Gyeonggi 16677 South Korea Seoul Natl Univ Dept Mech & Aerosp Engn Seoul 08826 South Korea Seoul Natl Univ Automat & Syst Res Inst Seoul 08826 South Korea Samsung Elect Samsung Res Seoul 497335 South Korea 

出 版 物：《IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING》 (IEEE Trans. Autom. Sci. Eng.)

年 卷 期：2022年第19卷第4期

页      面：3968-3979页

核心收录：

学科分类：0808[工学-电气工程] 08[工学] 0811[工学-控制科学与工程] 

基　　金：Ministry of Trade  Industry & Energy (MOTIE  South Korea) under Industrial Technology Innovation Program 

主　　题：Robots Trajectory Manipulator dynamics Heuristic algorithms Complexity theory Task analysis Real-time systems Motion-planning manipulator motion-planning mobile robot motion-planning robot programming multiple manipulators learning from demonstration motion representation algorithm parametric dynamic movement primitives dynamic movement primitives mobile manipulation 

摘      要：This paper presents an approach that generates the overall trajectory of mobile manipulators for a complex mission consisting of several sub-tasks. Parametric dynamic movement primitives (PDMPs) can quickly generalize the online motion of robot manipulation by learning multiple demonstrations in offline. However, regarding complex missions consisting of multiple sub-tasks, a large number of demonstrations are required for full generalization, which is impractical. In this paper, we propose a framework that reduces the number of demonstrations for a complex mission. In the proposed method, complex demonstrations are segmented into multiple unit motions representing sub-tasks, and one PDMP is formed per each segment, resulting in multiple PDMPs. The phase decision process determines which sub-task and associated PDMPs to be executed online, allowing multiple PDMPs to be autonomously configured within an integrated framework. In order to generalize the execution time and regional goal in each phase, the Gaussian process regression (GPR) is applied. Simulation results from two different scenarios confirm that the proposed framework not only effectively reduces the number of demonstrations but also improves generalization performance. The actual experiments also demonstrate that the mobile manipulators effectively perform complex missions through the proposed framework.