Robust Model Based Reinforcement Learning Using $\mathcal{L}_1$ Adaptive Control
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866916171988598784 |
|---|---|
| author | Sung, Minjun Karumanchi, Sambhu H. Gahlawat, Aditya Hovakimyan, Naira |
| author_facet | Sung, Minjun Karumanchi, Sambhu H. Gahlawat, Aditya Hovakimyan, Naira |
| contents | We introduce $\mathcal{L}_1$-MBRL, a control-theoretic augmentation scheme for Model-Based Reinforcement Learning (MBRL) algorithms. Unlike model-free approaches, MBRL algorithms learn a model of the transition function using data and use it to design a control input. Our approach generates a series of approximate control-affine models of the learned transition function according to the proposed switching law. Using the approximate model, control input produced by the underlying MBRL is perturbed by the $\mathcal{L}_1$ adaptive control, which is designed to enhance the robustness of the system against uncertainties. Importantly, this approach is agnostic to the choice of MBRL algorithm, enabling the use of the scheme with various MBRL algorithms. MBRL algorithms with $\mathcal{L}_1$ augmentation exhibit enhanced performance and sample efficiency across multiple MuJoCo environments, outperforming the original MBRL algorithms, both with and without system noise. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_14860 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Robust Model Based Reinforcement Learning Using $\mathcal{L}_1$ Adaptive Control Sung, Minjun Karumanchi, Sambhu H. Gahlawat, Aditya Hovakimyan, Naira Systems and Control Machine Learning We introduce $\mathcal{L}_1$-MBRL, a control-theoretic augmentation scheme for Model-Based Reinforcement Learning (MBRL) algorithms. Unlike model-free approaches, MBRL algorithms learn a model of the transition function using data and use it to design a control input. Our approach generates a series of approximate control-affine models of the learned transition function according to the proposed switching law. Using the approximate model, control input produced by the underlying MBRL is perturbed by the $\mathcal{L}_1$ adaptive control, which is designed to enhance the robustness of the system against uncertainties. Importantly, this approach is agnostic to the choice of MBRL algorithm, enabling the use of the scheme with various MBRL algorithms. MBRL algorithms with $\mathcal{L}_1$ augmentation exhibit enhanced performance and sample efficiency across multiple MuJoCo environments, outperforming the original MBRL algorithms, both with and without system noise. |
| title | Robust Model Based Reinforcement Learning Using $\mathcal{L}_1$ Adaptive Control |
| topic | Systems and Control Machine Learning |
| url | https://arxiv.org/abs/2403.14860 |