Title :
Learning reactive admittance control
Author :
Gullapalli, VijayKumar ; Grupen, Roderic A. ; Barto, Andrew G.
Author_Institution :
Dept. of Comput. Sci., Massachusetts Univ., Amherst, MA, USA
Abstract :
A peg-in-hole insertion task is used as an example to illustrate the utility of direct associative reinforcement learning methods for learning control under real-world conditions of uncertainty and noise. An associative reinforcement learning system has to learn appropriate actions in various situations through a search guided by evaluative performance feedback The authors used such a learning system, implemented as a connectionist network, to learn active compliant control for peg-in-hole insertion. The results indicated that direct reinforcement learning can be used to learn a reactive control strategy that works well even in the presence of a high degree of noise and uncertainty
Keywords :
assembling; learning (artificial intelligence); robots; search problems; direct associative reinforcement learning methods; evaluative performance feedback; learning; noise; peg-in-hole insertion task; reactive admittance control; search; uncertainty; Admittance; Computer science; Force control; Learning systems; Robot control; Robot sensing systems; Robotic assembly; Service robots; Strategic planning; Uncertainty;
Conference_Titel :
Robotics and Automation, 1992. Proceedings., 1992 IEEE International Conference on
Conference_Location :
Nice
Print_ISBN :
0-8186-2720-4
DOI :
10.1109/ROBOT.1992.220143