Physics-Informed Representation and Learning: Control and Risk Quantification
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2023
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914788853940224 |
|---|---|
| author | Wang, Zhuoyuan Keller, Reece Deng, Xiyu Hoshino, Kenta Tanaka, Takashi Nakahira, Yorie |
| author_facet | Wang, Zhuoyuan Keller, Reece Deng, Xiyu Hoshino, Kenta Tanaka, Takashi Nakahira, Yorie |
| contents | Optimal and safety-critical control are fundamental problems for stochastic systems, and are widely considered in real-world scenarios such as robotic manipulation and autonomous driving. In this paper, we consider the problem of efficiently finding optimal and safe control for high-dimensional systems. Specifically, we propose to use dimensionality reduction techniques from a comparison theorem for stochastic differential equations together with a generalizable physics-informed neural network to estimate the optimal value function and the safety probability of the system. The proposed framework results in substantial sample efficiency improvement compared to existing methods. We further develop an autoencoder-like neural network to automatically identify the low-dimensional features of the system to enhance the ease of design for system integration. We also provide experiments and quantitative analysis to validate the efficacy of the proposed method. Source code is available at https://github.com/jacobwang925/path-integral-PINN. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_10594 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Physics-Informed Representation and Learning: Control and Risk Quantification Wang, Zhuoyuan Keller, Reece Deng, Xiyu Hoshino, Kenta Tanaka, Takashi Nakahira, Yorie Systems and Control Optimal and safety-critical control are fundamental problems for stochastic systems, and are widely considered in real-world scenarios such as robotic manipulation and autonomous driving. In this paper, we consider the problem of efficiently finding optimal and safe control for high-dimensional systems. Specifically, we propose to use dimensionality reduction techniques from a comparison theorem for stochastic differential equations together with a generalizable physics-informed neural network to estimate the optimal value function and the safety probability of the system. The proposed framework results in substantial sample efficiency improvement compared to existing methods. We further develop an autoencoder-like neural network to automatically identify the low-dimensional features of the system to enhance the ease of design for system integration. We also provide experiments and quantitative analysis to validate the efficacy of the proposed method. Source code is available at https://github.com/jacobwang925/path-integral-PINN. |
| title | Physics-Informed Representation and Learning: Control and Risk Quantification |
| topic | Systems and Control |
| url | https://arxiv.org/abs/2312.10594 |