TY - GEN
T1 - Learning a Centroidal Motion Planner for Legged Locomotion
AU - Viereck, Julian
AU - Righetti, Ludovic
N1 - Funding Information:
This work was supported by the European Union's Horizon 2020 research and innovation program (grant agreement 780684 and European Research Council's grant 637935) and the National Science Foundation (grants 1825993, 1932187, 1925079 and 2026479).
Funding Information:
1 Tandon School of Engineering, New York University, USA jviereck@nyu.edu, ludovic.righetti@nyu.edu 2 Max Planck Institute for Intelligent Systems Tübingen, Germany This work was supported by the European Union’s Horizon 2020 research and innovation program (grant agreement 780684 and European Research Council’s grant 637935) and the National Science Foundation (grants 1825993, 1932187, 1925079 and 2026479).
Publisher Copyright:
© 2021 IEEE
PY - 2021
Y1 - 2021
N2 - Whole-body optimizers have been successful at automatically computing complex dynamic locomotion behaviors. However they are often limited to offline planning as they are computationally too expensive to replan with a high frequency. Simpler models are then typically used for online replanning. In this paper we present a method to generate whole body movements in real-time for locomotion tasks. Our approach consists in learning a centroidal neural network that predicts the desired centroidal motion given the current state of the robot and a desired contact plan. The network is trained using an existing whole body motion optimizer. Our approach enables to learn with few training samples dynamic motions that can be used in a complete whole-body control framework at high frequency, which is usually not attainable with typical full-body optimizers. We demonstrate our method to generate a rich set of walking and jumping motions on a real quadruped robot.
AB - Whole-body optimizers have been successful at automatically computing complex dynamic locomotion behaviors. However they are often limited to offline planning as they are computationally too expensive to replan with a high frequency. Simpler models are then typically used for online replanning. In this paper we present a method to generate whole body movements in real-time for locomotion tasks. Our approach consists in learning a centroidal neural network that predicts the desired centroidal motion given the current state of the robot and a desired contact plan. The network is trained using an existing whole body motion optimizer. Our approach enables to learn with few training samples dynamic motions that can be used in a complete whole-body control framework at high frequency, which is usually not attainable with typical full-body optimizers. We demonstrate our method to generate a rich set of walking and jumping motions on a real quadruped robot.
UR - http://www.scopus.com/inward/record.url?scp=85124348431&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85124348431&partnerID=8YFLogxK
U2 - 10.1109/ICRA48506.2021.9562022
DO - 10.1109/ICRA48506.2021.9562022
M3 - Conference contribution
AN - SCOPUS:85124348431
T3 - Proceedings - IEEE International Conference on Robotics and Automation
SP - 4905
EP - 4911
BT - 2021 IEEE International Conference on Robotics and Automation, ICRA 2021
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2021 IEEE International Conference on Robotics and Automation, ICRA 2021
Y2 - 30 May 2021 through 5 June 2021
ER -