Verfahren zum Trainieren eines Neuronalen Netzes zur Steuerung eines Flugzeugsystems

Details

Veröffentlichungs-Nr.

EP4052974

Aktenzeichen

EP22159778

Anmeldetag

2. März 2022

Veröffentlichung

21. Dezember 2022

Rechtsraum

EP

IPC

B60T8/174G06N3/08B64C13/50

Abstract

Disclosed is a method of training a neural network to control an aircraft system having a plurality of components. The method includes obtaining an operational mode data space representing a set of operational modes for the aircraft system, wherein each operational mode represents a configuration of the aircraft system where performance of one or more of the plurality of components is impaired according to a model of the aircraft system. Each operational mode comprises probability data indicating a respective probability of the operational mode occurring, according to the model of the aircraft system. The method includes generating a reduced operational mode data space comprising operational modes having a probability greater than a theoretical operational probability threshold, and generating a training operational mode data space comprising operational modes within the reduced operational mode data space that have a probability less than a real-world operational probability threshold. The method includes training the neural network using operational modes within the training operational mode data space.

Anmelder

Firmen

Airbus Operations Limited
Airbus S.A.S.

Land

🇬🇧 Großbritannien

Fachgebiete