Guardado en:
Detalles Bibliográficos
Autores principales: Lu, Yupu, Lin, Shijie, Xu, Hao, Zhang, Zeqing, Pan, Jia
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2511.01283
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912683644682240
author Lu, Yupu
Lin, Shijie
Xu, Hao
Zhang, Zeqing
Pan, Jia
author_facet Lu, Yupu
Lin, Shijie
Xu, Hao
Zhang, Zeqing
Pan, Jia
contents Finding a control Lyapunov function (CLF) in a dynamical system with a controller is an effective way to guarantee stability, which is a crucial issue in safety-concerned applications. Recently, deep learning models representing CLFs have been applied into a learner-verifier framework to identify satisfiable candidates. However, the learner treats Lyapunov conditions as complex constraints for optimisation, which is hard to achieve global convergence. It is also too complicated to implement these Lyapunov conditions for verification. To improve this framework, we treat Lyapunov conditions as inductive biases and design a neural CLF and a CLF-based controller guided by this knowledge. This design enables a stable optimisation process with limited constraints, and allows end-to-end learning of both the CLF and the controller. Our approach achieves a higher convergence rate and larger region of attraction (ROA) in learning the CLF compared to existing methods among abundant experiment cases. We also thoroughly reveal why the success rate decreases with previous methods during learning.
format Preprint
id arxiv_https___arxiv_org_abs_2511_01283
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Lyapunov Stability Learning with Nonlinear Control via Inductive Biases
Lu, Yupu
Lin, Shijie
Xu, Hao
Zhang, Zeqing
Pan, Jia
Machine Learning
Robotics
Finding a control Lyapunov function (CLF) in a dynamical system with a controller is an effective way to guarantee stability, which is a crucial issue in safety-concerned applications. Recently, deep learning models representing CLFs have been applied into a learner-verifier framework to identify satisfiable candidates. However, the learner treats Lyapunov conditions as complex constraints for optimisation, which is hard to achieve global convergence. It is also too complicated to implement these Lyapunov conditions for verification. To improve this framework, we treat Lyapunov conditions as inductive biases and design a neural CLF and a CLF-based controller guided by this knowledge. This design enables a stable optimisation process with limited constraints, and allows end-to-end learning of both the CLF and the controller. Our approach achieves a higher convergence rate and larger region of attraction (ROA) in learning the CLF compared to existing methods among abundant experiment cases. We also thoroughly reveal why the success rate decreases with previous methods during learning.
title Lyapunov Stability Learning with Nonlinear Control via Inductive Biases
topic Machine Learning
Robotics
url https://arxiv.org/abs/2511.01283