A Caltech Library Service

Neural Gaits: Learning Bipedal Locomotion via Control Barrier Functions and Zero Dynamics Policies

Jimenez Rodriguez, Ivan Dario and Csomay-Shanklin, Noel and Yue, Yisong and Ames, Aaron D. (2022) Neural Gaits: Learning Bipedal Locomotion via Control Barrier Functions and Zero Dynamics Policies. Proceedings of Machine Learning Research, 168 . pp. 1060-1072. ISSN 2640-3498.

[img] PDF - Published Version
See Usage Policy.


Use this Persistent URL to link to this item:


This work presents Neural Gaits, a method for learning dynamic walking gaits through the enforcement of set invariance that can be refined episodically using experimental data from the robot. We frame walking as a set invariance problem enforceable via control barrier functions (CBFs) defined on the reduced-order dynamics quantifying the underactuated component of the robot: the zero dynamics. Our approach contains two learning modules: one for learning a policy that satisfies the CBF condition, and another for learning a residual dynamics model to refine imperfections of the nominal model. Importantly, learning only over the zero dynamics significantly reduces the dimensionality of the learning problem while using CBFs allows us to still make guarantees for the full-order system. The method is demonstrated experimentally on an underactuated bipedal robot, where we are able to show agile and dynamic locomotion, even with partially unknown dynamics.

Item Type:Article
Related URLs:
URLURL TypeDescription Paper ItemLearning Code ItemVideo
Jimenez Rodriguez, Ivan Dario0000-0001-9065-5227
Csomay-Shanklin, Noel0000-0002-2361-1694
Yue, Yisong0000-0001-9127-1989
Ames, Aaron D.0000-0003-0848-3177
Additional Information:© 2022 I.D. Jimenez Rodriguez, N. Csomay-Shanklin, Y. Yue & A.D. Ames. The authors would like to thank Min Dai, Ryan Cosner, and Andrew Taylor for their insightful discussions related to walking, barrier functions, and projection to state safety.
Subject Keywords:bipedal locomotion, zero dynamics, safety, robotics
Record Number:CaltechAUTHORS:20220714-194300374
Persistent URL:
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:115562
Deposited By: George Porter
Deposited On:14 Jul 2022 22:53
Last Modified:14 Jul 2022 22:53

Repository Staff Only: item control page