TY - JOUR
T1 - Impedance learning for robotic contact tasks using natural actor-critic algorithm
AU - Kim, Byungchan
AU - Park, Jooyoung
AU - Park, Shinsuk
AU - Kang, Sungchul
N1 - Funding Information:
Manuscript received February 13, 2009; revised May 22, 2009. First published August 18, 2009; current version published March 17, 2010. This work was supported by the Korean Institute of Construction and Transportation Technology Evaluation and Planning under Program 06-Unified and Advanced Construction Technology Program-D01. The work of J. Park was supported by the Ministry of Knowledge Economy under the Human Resources Development Program for Convergence Robot Specialists. This paper was recommended by Associate Editor A. Tayebi.
Copyright:
Copyright 2010 Elsevier B.V., All rights reserved.
PY - 2010/4
Y1 - 2010/4
N2 - Compared with their robotic counterparts, humans excel at various tasks by using their ability to adaptively modulate arm impedance parameters. This ability allows us to successfully perform contact tasks even in uncertain environments. This paper considers a learning strategy of motor skill for robotic contact tasks based on a human motor control theory and machine learning schemes. Our robot learning method employs impedance control based on the equilibrium point control theory and reinforcement learning to determine the impedance parameters for contact tasks. A recursive least-square filter-based episodic natural actor-critic algorithm is used to find the optimal impedance parameters. The effectiveness of the proposed method was tested through dynamic simulations of various contact tasks. The simulation results demonstrated that the proposed method optimizes the performance of the contact tasks in uncertain conditions of the environment.
AB - Compared with their robotic counterparts, humans excel at various tasks by using their ability to adaptively modulate arm impedance parameters. This ability allows us to successfully perform contact tasks even in uncertain environments. This paper considers a learning strategy of motor skill for robotic contact tasks based on a human motor control theory and machine learning schemes. Our robot learning method employs impedance control based on the equilibrium point control theory and reinforcement learning to determine the impedance parameters for contact tasks. A recursive least-square filter-based episodic natural actor-critic algorithm is used to find the optimal impedance parameters. The effectiveness of the proposed method was tested through dynamic simulations of various contact tasks. The simulation results demonstrated that the proposed method optimizes the performance of the contact tasks in uncertain conditions of the environment.
KW - Contact task
KW - Equilibrium point control
KW - Reinforcement learning
KW - Robot manipulation
UR - http://www.scopus.com/inward/record.url?scp=77949776001&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=77949776001&partnerID=8YFLogxK
U2 - 10.1109/TSMCB.2009.2026289
DO - 10.1109/TSMCB.2009.2026289
M3 - Article
C2 - 19696001
AN - SCOPUS:77949776001
VL - 40
SP - 433
EP - 443
JO - IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
JF - IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
SN - 1083-4419
IS - 2
M1 - 5204203
ER -