Improving Input-Output Linearizing Controllers for Bipedal Robots via Reinforcement Learning

Fernando Castañeda, Mathias Wulfman, Ayush Agrawal, Tyler Westenbroek, Shankar Sastry, Claire Tomlin, Koushil Sreenath
Proceedings of the 2nd Conference on Learning for Dynamics and Control, PMLR 120:990-999, 2020.

Abstract

The need of precise dynamics models and not being able to account for input constraints are two of the main drawbacks of input-output linearizing controllers. Model uncertainty is common in almost every robotic application, and input saturation is present in every real world system. In this paper, we address both challenges for the specific case of bipedal robots’ control by the use of reinforcement learning techniques. We demonstrate the performance of the designed controller for different uncertain scenarios on the five-link planar robot RABBIT. The advantages of the designed controller are highlighted and a comparison with a known effective adaptive controller is presented.

Cite this Paper

BibTeX
@InProceedings{pmlr-v120-castaneda20a, title = {Improving Input-Output Linearizing Controllers for Bipedal Robots via Reinforcement Learning}, author = {Casta{\~n}eda, Fernando and Wulfman, Mathias and Agrawal, Ayush and Westenbroek, Tyler and Sastry, Shankar and Tomlin, Claire and Sreenath, Koushil}, booktitle = {Proceedings of the 2nd Conference on Learning for Dynamics and Control}, pages = {990--999}, year = {2020}, editor = {Bayen, Alexandre M. and Jadbabaie, Ali and Pappas, George and Parrilo, Pablo A. and Recht, Benjamin and Tomlin, Claire and Zeilinger, Melanie}, volume = {120}, series = {Proceedings of Machine Learning Research}, month = {10--11 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v120/castaneda20a/castaneda20a.pdf}, url = {https://proceedings.mlr.press/v120/castaneda20a.html}, abstract = {The need of precise dynamics models and not being able to account for input constraints are two of the main drawbacks of input-output linearizing controllers. Model uncertainty is common in almost every robotic application, and input saturation is present in every real world system. In this paper, we address both challenges for the specific case of bipedal robots’ control by the use of reinforcement learning techniques. We demonstrate the performance of the designed controller for different uncertain scenarios on the five-link planar robot RABBIT. The advantages of the designed controller are highlighted and a comparison with a known effective adaptive controller is presented.} }
Endnote
%0 Conference Paper %T Improving Input-Output Linearizing Controllers for Bipedal Robots via Reinforcement Learning %A Fernando Castañeda %A Mathias Wulfman %A Ayush Agrawal %A Tyler Westenbroek %A Shankar Sastry %A Claire Tomlin %A Koushil Sreenath %B Proceedings of the 2nd Conference on Learning for Dynamics and Control %C Proceedings of Machine Learning Research %D 2020 %E Alexandre M. Bayen %E Ali Jadbabaie %E George Pappas %E Pablo A. Parrilo %E Benjamin Recht %E Claire Tomlin %E Melanie Zeilinger %F pmlr-v120-castaneda20a %I PMLR %P 990--999 %U https://proceedings.mlr.press/v120/castaneda20a.html %V 120 %X The need of precise dynamics models and not being able to account for input constraints are two of the main drawbacks of input-output linearizing controllers. Model uncertainty is common in almost every robotic application, and input saturation is present in every real world system. In this paper, we address both challenges for the specific case of bipedal robots’ control by the use of reinforcement learning techniques. We demonstrate the performance of the designed controller for different uncertain scenarios on the five-link planar robot RABBIT. The advantages of the designed controller are highlighted and a comparison with a known effective adaptive controller is presented.
APA
Castañeda, F., Wulfman, M., Agrawal, A., Westenbroek, T., Sastry, S., Tomlin, C. & Sreenath, K.. (2020). Improving Input-Output Linearizing Controllers for Bipedal Robots via Reinforcement Learning. Proceedings of the 2nd Conference on Learning for Dynamics and Control, in Proceedings of Machine Learning Research 120:990-999 Available from https://proceedings.mlr.press/v120/castaneda20a.html.

Related Material