Reinforcement Learning for Robust Parameterized Locomotion Control of Bipedal Robots

IEEE International Conference on Robotics and Automation (ICRA 2021)

Zhongyu Li    Xuxin Cheng    Xue Bin Peng    Pieter Abbeel    Sergey Levine    Glen Berseth    Koushil Sreenath

University of California, Berkeley

---

Abstract

Developing robust walking controllers for bipedal robots is a challenging endeavor. Traditional model-based loco-motion controllers require simplifying assumptions and careful modelling; any small errors can result in unstable control. To address these challenges for bipedal locomotion, we presenta model-free reinforcement learning framework for training robust locomotion policies in simulation, which can then be transferred to a real bipedal Cassie robot. To facilitate sim-to-real transfer, domain randomization is used to encourage the policies to learn behaviors that are robust across variations in system dynamics. The learned policies enable Cassie to performa set of diverse and dynamic behaviors, while also being more robust than traditional controllers and prior learning-based methods that use residual control. We demonstrate this on versatile walking behaviors such as tracking a target walking velocity, walking height, and turning yaw.

Paper: [PDF]       Preprint: [arXiv]

Video

Bibtex

@inproceedings{
	CassieLi2021,
	author={Li, Zhongyu and Cheng, Xuxin and Peng, Xue Bin and Abbeel, Pieter and Levine, Sergey and Berseth, Glen and Sreenath, Koushil},
	booktitle={2021 IEEE International Conference on Robotics and Automation (ICRA)}, 
	title={Reinforcement Learning for Robust Parameterized Locomotion Control of Bipedal Robots}, 
	year={2021},
	volume={},
	number={},
	pages={2811-2817},
	doi={10.1109/ICRA48506.2021.9560769}
}