Adaptive Model Predictive Control by Learning Classifiers

Rel Guzman, Rafael Oliveira, Fabio Ramos
Proceedings of The 4th Annual Learning for Dynamics and Control Conference, PMLR 168:480-491, 2022.

Abstract

Stochastic model predictive control has been a successful and robust control framework for many robotics tasks where the system dynamics model is slightly inaccurate or in the presence of environment disturbances. Despite the successes, it is still unclear how to best adjust control parameters to the current task in the presence of model parameter uncertainty and heteroscedastic noise. In this paper, we propose an adaptive MPC variant that automatically estimates control and model parameters by leveraging ideas from Bayesian optimisation (BO) and the classical expected improvement acquisition function. We leverage recent results showing that BO can be reformulated via density ratio estimation, which can be efficiently approximated by simply learning a classifier. This is then integrated into a model predictive path integral control framework yielding robust controllers for a variety of challenging robotics tasks. We demonstrate the approach on classical control problems under model uncertainty and robotics manipulation tasks.

Cite this Paper


BibTeX
@InProceedings{pmlr-v168-guzman22a, title = {Adaptive Model Predictive Control by Learning Classifiers}, author = {Guzman, Rel and Oliveira, Rafael and Ramos, Fabio}, booktitle = {Proceedings of The 4th Annual Learning for Dynamics and Control Conference}, pages = {480--491}, 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/guzman22a/guzman22a.pdf}, url = {https://proceedings.mlr.press/v168/guzman22a.html}, abstract = {Stochastic model predictive control has been a successful and robust control framework for many robotics tasks where the system dynamics model is slightly inaccurate or in the presence of environment disturbances. Despite the successes, it is still unclear how to best adjust control parameters to the current task in the presence of model parameter uncertainty and heteroscedastic noise. In this paper, we propose an adaptive MPC variant that automatically estimates control and model parameters by leveraging ideas from Bayesian optimisation (BO) and the classical expected improvement acquisition function. We leverage recent results showing that BO can be reformulated via density ratio estimation, which can be efficiently approximated by simply learning a classifier. This is then integrated into a model predictive path integral control framework yielding robust controllers for a variety of challenging robotics tasks. We demonstrate the approach on classical control problems under model uncertainty and robotics manipulation tasks.} }
Endnote
%0 Conference Paper %T Adaptive Model Predictive Control by Learning Classifiers %A Rel Guzman %A Rafael Oliveira %A Fabio Ramos %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-guzman22a %I PMLR %P 480--491 %U https://proceedings.mlr.press/v168/guzman22a.html %V 168 %X Stochastic model predictive control has been a successful and robust control framework for many robotics tasks where the system dynamics model is slightly inaccurate or in the presence of environment disturbances. Despite the successes, it is still unclear how to best adjust control parameters to the current task in the presence of model parameter uncertainty and heteroscedastic noise. In this paper, we propose an adaptive MPC variant that automatically estimates control and model parameters by leveraging ideas from Bayesian optimisation (BO) and the classical expected improvement acquisition function. We leverage recent results showing that BO can be reformulated via density ratio estimation, which can be efficiently approximated by simply learning a classifier. This is then integrated into a model predictive path integral control framework yielding robust controllers for a variety of challenging robotics tasks. We demonstrate the approach on classical control problems under model uncertainty and robotics manipulation tasks.
APA
Guzman, R., Oliveira, R. & Ramos, F.. (2022). Adaptive Model Predictive Control by Learning Classifiers. Proceedings of The 4th Annual Learning for Dynamics and Control Conference, in Proceedings of Machine Learning Research 168:480-491 Available from https://proceedings.mlr.press/v168/guzman22a.html.

Related Material