Asynchronous Deep Model Reference Adaptive Control

Girish Joshi, Jasvir Virdi, Girish Chowdhary
Proceedings of the 2020 Conference on Robot Learning, PMLR 155:984-1000, 2021.

Abstract

In this paper, we present Asynchronous implementation of Deep Neural Network-based Model Reference Adaptive Control (DMRAC). We evaluate this new neuro-adaptive control architecture through flight tests on a small quadcopter. We demonstrate that a single DMRAC controller can handle significant nonlinearities due to severe system faults and deliberate wind disturbances while executing high-bandwidth attitude control. We also show that the architecture has long-term learning abilities across different flight regimes, and can generalize to fly different flight trajectories than those on which it was trained. These results demonstrating the efficacy of this architecture for high bandwidth closed-loop attitude control of unstable and nonlinear robots operating in adverse situations. To achieve these results, we designed a software+communication architecture to ensure online real-time inference of the deep network on a high-bandwidth computation-limited platform. We expect that this architecture will benefit other deep learning in the closed-loop experiments on robots.

Cite this Paper

BibTeX
@InProceedings{pmlr-v155-joshi21a, title = {Asynchronous Deep Model Reference Adaptive Control}, author = {Joshi, Girish and Virdi, Jasvir and Chowdhary, Girish}, booktitle = {Proceedings of the 2020 Conference on Robot Learning}, pages = {984--1000}, year = {2021}, editor = {Kober, Jens and Ramos, Fabio and Tomlin, Claire}, volume = {155}, series = {Proceedings of Machine Learning Research}, month = {16--18 Nov}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v155/joshi21a/joshi21a.pdf}, url = {https://proceedings.mlr.press/v155/joshi21a.html}, abstract = {In this paper, we present Asynchronous implementation of Deep Neural Network-based Model Reference Adaptive Control (DMRAC). We evaluate this new neuro-adaptive control architecture through flight tests on a small quadcopter. We demonstrate that a single DMRAC controller can handle significant nonlinearities due to severe system faults and deliberate wind disturbances while executing high-bandwidth attitude control. We also show that the architecture has long-term learning abilities across different flight regimes, and can generalize to fly different flight trajectories than those on which it was trained. These results demonstrating the efficacy of this architecture for high bandwidth closed-loop attitude control of unstable and nonlinear robots operating in adverse situations. To achieve these results, we designed a software+communication architecture to ensure online real-time inference of the deep network on a high-bandwidth computation-limited platform. We expect that this architecture will benefit other deep learning in the closed-loop experiments on robots.} }
Endnote
%0 Conference Paper %T Asynchronous Deep Model Reference Adaptive Control %A Girish Joshi %A Jasvir Virdi %A Girish Chowdhary %B Proceedings of the 2020 Conference on Robot Learning %C Proceedings of Machine Learning Research %D 2021 %E Jens Kober %E Fabio Ramos %E Claire Tomlin %F pmlr-v155-joshi21a %I PMLR %P 984--1000 %U https://proceedings.mlr.press/v155/joshi21a.html %V 155 %X In this paper, we present Asynchronous implementation of Deep Neural Network-based Model Reference Adaptive Control (DMRAC). We evaluate this new neuro-adaptive control architecture through flight tests on a small quadcopter. We demonstrate that a single DMRAC controller can handle significant nonlinearities due to severe system faults and deliberate wind disturbances while executing high-bandwidth attitude control. We also show that the architecture has long-term learning abilities across different flight regimes, and can generalize to fly different flight trajectories than those on which it was trained. These results demonstrating the efficacy of this architecture for high bandwidth closed-loop attitude control of unstable and nonlinear robots operating in adverse situations. To achieve these results, we designed a software+communication architecture to ensure online real-time inference of the deep network on a high-bandwidth computation-limited platform. We expect that this architecture will benefit other deep learning in the closed-loop experiments on robots.
APA
Joshi, G., Virdi, J. & Chowdhary, G.. (2021). Asynchronous Deep Model Reference Adaptive Control. Proceedings of the 2020 Conference on Robot Learning, in Proceedings of Machine Learning Research 155:984-1000 Available from https://proceedings.mlr.press/v155/joshi21a.html.

Related Material