Learning Performance-Oriented Control Barrier Functions Under Complex Safety Constraints and Limited Actuation

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Manda, Lakshmideepakreddy, Chen, Shaoru, Fazlyab, Mahyar
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866910680942116864
author Manda, Lakshmideepakreddy
Chen, Shaoru
Fazlyab, Mahyar
author_facet Manda, Lakshmideepakreddy
Chen, Shaoru
Fazlyab, Mahyar
contents Control Barrier Functions (CBFs) provide an elegant framework for constraining nonlinear control system dynamics to remain within an invariant subset of a designated safe set. However, identifying a CBF that balances performance-by maximizing the control invariant set-and accommodates complex safety constraints, especially in systems with high relative degree and actuation limits, poses a significant challenge. In this work, we introduce a novel self-supervised learning framework to comprehensively address these challenges. Our method begins with a Boolean composition of multiple state constraints that define the safe set. We first construct a smooth function whose zero superlevel set forms an inner approximation of this safe set. This function is then combined with a smooth neural network to parameterize the CBF candidate. To train the CBF and maximize the volume of the resulting control invariant set, we design a physics-informed loss function based on a Hamilton-Jacobi Partial Differential Equation (PDE). We validate the efficacy of our approach on a 2D double integrator (DI) system and a 7D fixed-wing aircraft system (F16).
format Preprint
id arxiv_https___arxiv_org_abs_2401_05629
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Learning Performance-Oriented Control Barrier Functions Under Complex Safety Constraints and Limited Actuation
Manda, Lakshmideepakreddy
Chen, Shaoru
Fazlyab, Mahyar
Machine Learning
Systems and Control
Control Barrier Functions (CBFs) provide an elegant framework for constraining nonlinear control system dynamics to remain within an invariant subset of a designated safe set. However, identifying a CBF that balances performance-by maximizing the control invariant set-and accommodates complex safety constraints, especially in systems with high relative degree and actuation limits, poses a significant challenge. In this work, we introduce a novel self-supervised learning framework to comprehensively address these challenges. Our method begins with a Boolean composition of multiple state constraints that define the safe set. We first construct a smooth function whose zero superlevel set forms an inner approximation of this safe set. This function is then combined with a smooth neural network to parameterize the CBF candidate. To train the CBF and maximize the volume of the resulting control invariant set, we design a physics-informed loss function based on a Hamilton-Jacobi Partial Differential Equation (PDE). We validate the efficacy of our approach on a 2D double integrator (DI) system and a 7D fixed-wing aircraft system (F16).
title Learning Performance-Oriented Control Barrier Functions Under Complex Safety Constraints and Limited Actuation
topic Machine Learning
Systems and Control
url https://arxiv.org/abs/2401.05629