Knowledge Network Node
Journal of Computer Applications. Core
(Accepted Version) Online First Publishing Date: 2023-05-26 11:11:30

Gait control method based on maximum entropy deep reinforcement learning for biped robotChinese Full Text

LI Yuanchao;TAO Chongben;WANG Chen;

Abstract: For the problem that gait stability control for continuous linear walking of a biped robot, a soft actor-critic（SAC） gait control algorithm based on maximum entropy deep reinforcement learning was proposed. Firstly, an accurate robot dynamics model did not need to be built in advance, all parameters were derived from joint angles without additional sensors. Secondly, the cosine similarity method was introduced to classify experience samples and optimize the experience replay mechanism. Finally, ... More
Keywords：

biped robot; gait control; deep reinforcement learning; maximum entropy; soft actor-critic algorithm;

  • Series：

    (I) Electronic Technology ＆ Information Science

  • Subject：

    Automation Technology

  • Classification Code：

    TP242;TP273

CNKI exclusive online-first articles is prohibited to reprint and excerpt without permission.

Download the mobile appuse the app to scan this coderead the article.

Tips: Please download CAJViewer to view CAJ format full text.

Download: 237 Pagecount: 7 Size: 841K

Related Literature
  • Similar Article
  • Reader Recommendation
  • Associated Author