L1-GP: L1 Adaptive Control with Bayesian Learning

Aditya Gahlawat, Pan Zhao, Andrew Patterson, Naira Hovakimyan, Evangelos Theodorou
Proceedings of the 2nd Conference on Learning for Dynamics and Control, PMLR 120:826-837, 2020.

Abstract

We present L1-GP, an architecture based on L1 adaptive control and Gaussian Process Regression (GPR) for safe simultaneous control and learning. On one hand, the L1 adaptive control provides stability and transient performance guarantees, which allows for GPR to efficiently and safely learn the uncertain dynamics. On the other hand, the learned dynamics can be conveniently incorporated into the L1 control architecture without sacrificing robustness and tracking performance. Subsequently, the learned dynamics can lead to less conservative designs for performance/robustness tradeoff. We illustrate the efficacy of the proposed architecture via numerical simulations.

Cite this Paper

BibTeX
@InProceedings{pmlr-v120-gahlawat20a, title = {L1-GP: L1 Adaptive Control with Bayesian Learning}, author = {Gahlawat, Aditya and Zhao, Pan and Patterson, Andrew and Hovakimyan, Naira and Theodorou, Evangelos}, booktitle = {Proceedings of the 2nd Conference on Learning for Dynamics and Control}, pages = {826--837}, 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/gahlawat20a/gahlawat20a.pdf}, url = {https://proceedings.mlr.press/v120/gahlawat20a.html}, abstract = {We present L1-GP, an architecture based on L1 adaptive control and Gaussian Process Regression (GPR) for safe simultaneous control and learning. On one hand, the L1 adaptive control provides stability and transient performance guarantees, which allows for GPR to efficiently and safely learn the uncertain dynamics. On the other hand, the learned dynamics can be conveniently incorporated into the L1 control architecture without sacrificing robustness and tracking performance. Subsequently, the learned dynamics can lead to less conservative designs for performance/robustness tradeoff. We illustrate the efficacy of the proposed architecture via numerical simulations.} }
Endnote
%0 Conference Paper %T L1-GP: L1 Adaptive Control with Bayesian Learning %A Aditya Gahlawat %A Pan Zhao %A Andrew Patterson %A Naira Hovakimyan %A Evangelos Theodorou %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-gahlawat20a %I PMLR %P 826--837 %U https://proceedings.mlr.press/v120/gahlawat20a.html %V 120 %X We present L1-GP, an architecture based on L1 adaptive control and Gaussian Process Regression (GPR) for safe simultaneous control and learning. On one hand, the L1 adaptive control provides stability and transient performance guarantees, which allows for GPR to efficiently and safely learn the uncertain dynamics. On the other hand, the learned dynamics can be conveniently incorporated into the L1 control architecture without sacrificing robustness and tracking performance. Subsequently, the learned dynamics can lead to less conservative designs for performance/robustness tradeoff. We illustrate the efficacy of the proposed architecture via numerical simulations.
APA
Gahlawat, A., Zhao, P., Patterson, A., Hovakimyan, N. & Theodorou, E.. (2020). L1-GP: L1 Adaptive Control with Bayesian Learning. Proceedings of the 2nd Conference on Learning for Dynamics and Control, in Proceedings of Machine Learning Research 120:826-837 Available from https://proceedings.mlr.press/v120/gahlawat20a.html.

Related Material