Multi-Task Reinforcement Learning based Mobile Manipulation Control for Dynamic Object Tracking and Grasping

Cong Wang, Qifeng Zhang, Xiaohui Wang, Shida Xu, Yvan Petillot, Sen Wang

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

3 Citations (Scopus)

Abstract

Agile control of mobile manipulator is challenging because of the high complexity coupled by the robotic system and the unstructured working environment. Tracking and grasping a dynamic object with a random trajectory is even harder. In this paper, a multi-task reinforcement learning-based mobile manipulation control framework is proposed to achieve general dynamic object tracking and grasping. Several basic types of dynamic trajectories are chosen as the training set for the task. To improve policy generalization in practice, random noise and dynamics randomization are introduced during the training process. Extensive experiments show that our trained policy can adapt to unseen random dynamic trajectories with about 0.1 m tracking error and 75% grasping success rate for dynamic objects. The trained policy can also be successfully deployed on a real mobile manipulator.

Original languageEnglish
Title of host publication7th Asia-Pacific Conference on Intelligent Robot Systems 2022
PublisherIEEE
Pages34-40
Number of pages7
ISBN (Electronic)9781665485197
DOIs
Publication statusPublished - 18 Aug 2022
Event7th Asia-Pacific Conference on Intelligent Robot Systems 2022 - Tianjin, China
Duration: 1 Jul 20223 Jul 2022

Conference

Conference7th Asia-Pacific Conference on Intelligent Robot Systems 2022
Abbreviated titleACIRS 2022
Country/TerritoryChina
CityTianjin
Period1/07/223/07/22

Keywords

  • Dynamic Object
  • Mobile Manipulation
  • Reinforcement Learning

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Mechanical Engineering
  • Control and Optimization

Fingerprint

Dive into the research topics of 'Multi-Task Reinforcement Learning based Mobile Manipulation Control for Dynamic Object Tracking and Grasping'. Together they form a unique fingerprint.

Cite this