RL + Model-based Control: Using On-demand Optimal Control to Learn Versatile Legged Locomotion

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kang, Dongho, Cheng, Jin, Zamora, Miguel, Zargarbashi, Fatemeh, Coros, Stelian
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912048634396672
author Kang, Dongho
Cheng, Jin
Zamora, Miguel
Zargarbashi, Fatemeh
Coros, Stelian
author_facet Kang, Dongho
Cheng, Jin
Zamora, Miguel
Zargarbashi, Fatemeh
Coros, Stelian
contents This paper presents a control framework that combines model-based optimal control and reinforcement learning (RL) to achieve versatile and robust legged locomotion. Our approach enhances the RL training process by incorporating on-demand reference motions generated through finite-horizon optimal control, covering a broad range of velocities and gaits. These reference motions serve as targets for the RL policy to imitate, leading to the development of robust control policies that can be learned with reliability. Furthermore, by utilizing realistic simulation data that captures whole-body dynamics, RL effectively overcomes the inherent limitations in reference motions imposed by modeling simplifications. We validate the robustness and controllability of the RL training process within our framework through a series of experiments. In these experiments, our method showcases its capability to generalize reference motions and effectively handle more complex locomotion tasks that may pose challenges for the simplified model, thanks to RL's flexibility. Additionally, our framework effortlessly supports the training of control policies for robots with diverse dimensions, eliminating the necessity for robot-specific adjustments in the reward function and hyperparameters.
format Preprint
id arxiv_https___arxiv_org_abs_2305_17842
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle RL + Model-based Control: Using On-demand Optimal Control to Learn Versatile Legged Locomotion
Kang, Dongho
Cheng, Jin
Zamora, Miguel
Zargarbashi, Fatemeh
Coros, Stelian
Robotics
Artificial Intelligence
Systems and Control
This paper presents a control framework that combines model-based optimal control and reinforcement learning (RL) to achieve versatile and robust legged locomotion. Our approach enhances the RL training process by incorporating on-demand reference motions generated through finite-horizon optimal control, covering a broad range of velocities and gaits. These reference motions serve as targets for the RL policy to imitate, leading to the development of robust control policies that can be learned with reliability. Furthermore, by utilizing realistic simulation data that captures whole-body dynamics, RL effectively overcomes the inherent limitations in reference motions imposed by modeling simplifications. We validate the robustness and controllability of the RL training process within our framework through a series of experiments. In these experiments, our method showcases its capability to generalize reference motions and effectively handle more complex locomotion tasks that may pose challenges for the simplified model, thanks to RL's flexibility. Additionally, our framework effortlessly supports the training of control policies for robots with diverse dimensions, eliminating the necessity for robot-specific adjustments in the reward function and hyperparameters.
title RL + Model-based Control: Using On-demand Optimal Control to Learn Versatile Legged Locomotion
topic Robotics
Artificial Intelligence
Systems and Control
url https://arxiv.org/abs/2305.17842