System-level Safety Guard: Safe Tracking Control through Uncertain Neural Network Dynamics Models

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Li, Xiao, Li, Yutong, Girard, Anouck, Kolmanovsky, Ilya
Format: Preprint
Veröffentlicht: 2023
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866913355474665472
author Li, Xiao
Li, Yutong
Girard, Anouck
Kolmanovsky, Ilya
author_facet Li, Xiao
Li, Yutong
Girard, Anouck
Kolmanovsky, Ilya
contents The Neural Network (NN), as a black-box function approximator, has been considered in many control and robotics applications. However, difficulties in verifying the overall system safety in the presence of uncertainties hinder the deployment of NN modules in safety-critical systems. In this paper, we leverage the NNs as predictive models for trajectory tracking of unknown dynamical systems. We consider controller design in the presence of both intrinsic uncertainty and uncertainties from other system modules. In this setting, we formulate the constrained trajectory tracking problem and show that it can be solved using Mixed-integer Linear Programming (MILP). The proposed MILP-based approach is empirically demonstrated in robot navigation and obstacle avoidance through simulations. The demonstration videos are available at https://xiaolisean.github.io/publication/2023-11-01-L4DC2024.
format Preprint
id arxiv_https___arxiv_org_abs_2312_06810
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle System-level Safety Guard: Safe Tracking Control through Uncertain Neural Network Dynamics Models
Li, Xiao
Li, Yutong
Girard, Anouck
Kolmanovsky, Ilya
Robotics
Machine Learning
Systems and Control
Optimization and Control
The Neural Network (NN), as a black-box function approximator, has been considered in many control and robotics applications. However, difficulties in verifying the overall system safety in the presence of uncertainties hinder the deployment of NN modules in safety-critical systems. In this paper, we leverage the NNs as predictive models for trajectory tracking of unknown dynamical systems. We consider controller design in the presence of both intrinsic uncertainty and uncertainties from other system modules. In this setting, we formulate the constrained trajectory tracking problem and show that it can be solved using Mixed-integer Linear Programming (MILP). The proposed MILP-based approach is empirically demonstrated in robot navigation and obstacle avoidance through simulations. The demonstration videos are available at https://xiaolisean.github.io/publication/2023-11-01-L4DC2024.
title System-level Safety Guard: Safe Tracking Control through Uncertain Neural Network Dynamics Models
topic Robotics
Machine Learning
Systems and Control
Optimization and Control
url https://arxiv.org/abs/2312.06810