Learning Neural Network Controllers with Certified Robust Performance via Adversarial Training

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Junnarkar, Neelay, Sonmez, Yasin, Arcak, Murat
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910094874116096
author Junnarkar, Neelay
Sonmez, Yasin
Arcak, Murat
author_facet Junnarkar, Neelay
Sonmez, Yasin
Arcak, Murat
contents Neural network (NN) controllers achieve strong empirical performance on nonlinear dynamical systems, yet deploying them in safety-critical settings requires robustness to disturbances and uncertainty. We present a method for jointly synthesizing NN controllers and dissipativity certificates that formally guarantee robust closed-loop performance using adversarial training, in which we use counterexamples to the robust dissipativity condition to guide training. Verification is done post-training using alpha,beta-CROWN, a branch-and-bound-based method that enables direct analysis of the nonlinear dynamical system. The proposed method uses quadratic constraints (QCs) only for characterization of non-parametric uncertainties. The method is tested in numerical experiments on maximizing the volume of the set on which a system is certified to be robustly dissipative. Our method certifies regions up to 78 times larger than the region certified by a linear matrix inequality-based approach that we derive for comparison.
format Preprint
id arxiv_https___arxiv_org_abs_2604_01188
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Learning Neural Network Controllers with Certified Robust Performance via Adversarial Training
Junnarkar, Neelay
Sonmez, Yasin
Arcak, Murat
Systems and Control
Neural network (NN) controllers achieve strong empirical performance on nonlinear dynamical systems, yet deploying them in safety-critical settings requires robustness to disturbances and uncertainty. We present a method for jointly synthesizing NN controllers and dissipativity certificates that formally guarantee robust closed-loop performance using adversarial training, in which we use counterexamples to the robust dissipativity condition to guide training. Verification is done post-training using alpha,beta-CROWN, a branch-and-bound-based method that enables direct analysis of the nonlinear dynamical system. The proposed method uses quadratic constraints (QCs) only for characterization of non-parametric uncertainties. The method is tested in numerical experiments on maximizing the volume of the set on which a system is certified to be robustly dissipative. Our method certifies regions up to 78 times larger than the region certified by a linear matrix inequality-based approach that we derive for comparison.
title Learning Neural Network Controllers with Certified Robust Performance via Adversarial Training
topic Systems and Control
url https://arxiv.org/abs/2604.01188