An experience-based policy gradient method for smooth manipulation

  • Yongchao Wang
  • , Xuguang Lan
  • , Chuzhen Feng
  • , Lipeng Wan
  • , Jin Li
  • , Yuwang Liu
  • , Decai Li

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

4 Scopus citations

Abstract

Policy gradient methods have achieved remarkable success in continuous controlling tasks. However, in robotic control, original policy gradient algorithms depend on the first succeed experience which is usually a suboptimal solution. To improve the performance, we propose an experience-based policy gradient method(EBDDPG) which guides the robot to move in a smooth way. Besides, extra OU-noise is added to the action space to improve exploration. We tested our algorithm on Gazebo simulation environment with Baxter robot. The experimental results show our method guides the robot to manipulate more smoothly and improves success rate of grasping tasks.

Original languageEnglish
Title of host publication9th IEEE International Conference on Cyber Technology in Automation, Control and Intelligent Systems, CYBER 2019
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages93-97
Number of pages5
ISBN (Electronic)9781728107691
DOIs
StatePublished - Jul 2019
Event9th IEEE International Conference on Cyber Technology in Automation, Control and Intelligent Systems, CYBER 2019 - Suzhou, China
Duration: 29 Jul 20192 Aug 2019

Publication series

Name9th IEEE International Conference on Cyber Technology in Automation, Control and Intelligent Systems, CYBER 2019

Conference

Conference9th IEEE International Conference on Cyber Technology in Automation, Control and Intelligent Systems, CYBER 2019
Country/TerritoryChina
CitySuzhou
Period29/07/192/08/19

Keywords

  • Deep Reinforcement Learning
  • Policy Gradient
  • Robot Manipulation

Fingerprint

Dive into the research topics of 'An experience-based policy gradient method for smooth manipulation'. Together they form a unique fingerprint.

Cite this