Barrier Certificates (BCs) provide safety guarantees for dynamical systems. In this work, we attempt to synthesize a neural network BC for a collision avoidance system for a remotely controlled aircraft. We use a counterexample-guided (CEGIS) procedure composed of two elements: a learner that synthesizes a candidate BC, and a sound verifier that either certifies the candidate's validity or generates counterexamples to further guide the learner.

Barrier Certificates (BCs) provide safety guarantees for dynamical systems. In this work, we attempt to synthesize a neural network BC for a collision avoidance system for a remotely controlled aircraft. We use a counterexample-guided (CEGIS) procedure composed of two elements: a learner that synthesizes a candidate BC, and a sound verifier that either certifies the candidate's validity or generates counterexamples to further guide the learner.

Barrier Certificate Synthesis for Airborne Collision Avoidance

VERONESE, LUCA
2025/2026

Abstract

Barrier Certificates (BCs) provide safety guarantees for dynamical systems. In this work, we attempt to synthesize a neural network BC for a collision avoidance system for a remotely controlled aircraft. We use a counterexample-guided (CEGIS) procedure composed of two elements: a learner that synthesizes a candidate BC, and a sound verifier that either certifies the candidate's validity or generates counterexamples to further guide the learner.
2025
Barrier Certificate Synthesis for Airborne Collision Avoidance
Barrier Certificates (BCs) provide safety guarantees for dynamical systems. In this work, we attempt to synthesize a neural network BC for a collision avoidance system for a remotely controlled aircraft. We use a counterexample-guided (CEGIS) procedure composed of two elements: a learner that synthesizes a candidate BC, and a sound verifier that either certifies the candidate's validity or generates counterexamples to further guide the learner.
Barrier Certificate
NN controlled system
CEGIS
Collision avoidance
Unmanned aircraft
File in questo prodotto:
File Dimensione Formato  
veronese_luca.pdf

Accesso riservato

Dimensione 2.38 MB
Formato Adobe PDF
2.38 MB Adobe PDF

The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12608/108175