Learning Task-Specific Dynamics to Improve Whole-Body Control

Andrej Gams, Sean A. Mason, Ales Ude, Stefan Schaal, Ludovic Rigbetti

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In task-based inverse dynamics control, reference accelerations used to follow a desired plan can be broken down into feedforward and feedback trajectories. The feedback term accounts for tracking errors that are caused from inaccurate dynamic models or external disturbances. On underactuated, free-floating robots, such as humanoids, good tracking accuracy often necessitates high feedback gains, which leads to undesirable stiff behaviors. The magnitude of these gains is anyways often strongly limited by the control bandwidth. In this paper, we show how to reduce the required contribution of the feedback controller by incorporating learned task-space reference accelerations. Thus, we i) improve the execution of the given specific task, and ii) offer the means to reduce feedback gains, providing for greater compliance of the system. In contrast to learning task-specific joint-torques, which might produce a similar effect but can lead to poor generalization, our approach directly learns the task-space dynamics of the center of mass of a humanoid robot. Simulated and real-world results on the lower part of the Sarcos Hermes humanoid robot demonstrate the applicability of the approach.

Original languageEnglish (US)
Title of host publication2018 IEEE-RAS 18th International Conference on Humanoid Robots, Humanoids 2018
PublisherIEEE Computer Society
Pages658-663
Number of pages6
ISBN (Electronic)9781538672839
DOIs
StatePublished - Jul 2 2018
Event18th IEEE-RAS International Conference on Humanoid Robots, Humanoids 2018 - Beijing, China
Duration: Nov 6 2018Nov 9 2018

Publication series

NameIEEE-RAS International Conference on Humanoid Robots
Volume2018-November
ISSN (Print)2164-0572
ISSN (Electronic)2164-0580

Conference

Conference18th IEEE-RAS International Conference on Humanoid Robots, Humanoids 2018
Country/TerritoryChina
CityBeijing
Period11/6/1811/9/18

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Hardware and Architecture
  • Human-Computer Interaction
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Learning Task-Specific Dynamics to Improve Whole-Body Control'. Together they form a unique fingerprint.

Cite this