ValueNetQP: Learned One-step Optimal Control for Legged Locomotion

Julian Viereck, Avadesh Meduri, Ludovic Righetti
Proceedings of The 4th Annual Learning for Dynamics and Control Conference, PMLR 168:931-942, 2022.

Abstract

Optimal control is a successful approach to generate motions for complex robots, in particular for legged locomotion. However, these techniques are often too slow to run in real time for model predictive control or one needs to drastically simplify the dynamics model. In this work, we present a method to learn to predict the gradient and hessian of the problem value function, enabling fast resolution of the predictive control problem with a one-step quadratic program. In addition, our method is able to satisfy constraints like friction cones and unilateral constraints, which are important for high dynamics locomotion tasks. We demonstrate the capability of our method in simulation and on a real quadruped robot performing trotting and bounding motions.

Cite this Paper

BibTeX
@InProceedings{pmlr-v168-viereck22a, title = {ValueNetQP: Learned One-step Optimal Control for Legged Locomotion}, author = {Viereck, Julian and Meduri, Avadesh and Righetti, Ludovic}, booktitle = {Proceedings of The 4th Annual Learning for Dynamics and Control Conference}, pages = {931--942}, year = {2022}, editor = {Firoozi, Roya and Mehr, Negar and Yel, Esen and Antonova, Rika and Bohg, Jeannette and Schwager, Mac and Kochenderfer, Mykel}, volume = {168}, series = {Proceedings of Machine Learning Research}, month = {23--24 Jun}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v168/viereck22a/viereck22a.pdf}, url = {https://proceedings.mlr.press/v168/viereck22a.html}, abstract = {Optimal control is a successful approach to generate motions for complex robots, in particular for legged locomotion. However, these techniques are often too slow to run in real time for model predictive control or one needs to drastically simplify the dynamics model. In this work, we present a method to learn to predict the gradient and hessian of the problem value function, enabling fast resolution of the predictive control problem with a one-step quadratic program. In addition, our method is able to satisfy constraints like friction cones and unilateral constraints, which are important for high dynamics locomotion tasks. We demonstrate the capability of our method in simulation and on a real quadruped robot performing trotting and bounding motions.} }
Endnote
%0 Conference Paper %T ValueNetQP: Learned One-step Optimal Control for Legged Locomotion %A Julian Viereck %A Avadesh Meduri %A Ludovic Righetti %B Proceedings of The 4th Annual Learning for Dynamics and Control Conference %C Proceedings of Machine Learning Research %D 2022 %E Roya Firoozi %E Negar Mehr %E Esen Yel %E Rika Antonova %E Jeannette Bohg %E Mac Schwager %E Mykel Kochenderfer %F pmlr-v168-viereck22a %I PMLR %P 931--942 %U https://proceedings.mlr.press/v168/viereck22a.html %V 168 %X Optimal control is a successful approach to generate motions for complex robots, in particular for legged locomotion. However, these techniques are often too slow to run in real time for model predictive control or one needs to drastically simplify the dynamics model. In this work, we present a method to learn to predict the gradient and hessian of the problem value function, enabling fast resolution of the predictive control problem with a one-step quadratic program. In addition, our method is able to satisfy constraints like friction cones and unilateral constraints, which are important for high dynamics locomotion tasks. We demonstrate the capability of our method in simulation and on a real quadruped robot performing trotting and bounding motions.
APA
Viereck, J., Meduri, A. & Righetti, L.. (2022). ValueNetQP: Learned One-step Optimal Control for Legged Locomotion. Proceedings of The 4th Annual Learning for Dynamics and Control Conference, in Proceedings of Machine Learning Research 168:931-942 Available from https://proceedings.mlr.press/v168/viereck22a.html.

Related Material