System und Verfahren zur Bedingten Prädiktiven Wartung unter Verwendung von Surrogatmodellierung

Details

Veröffentlichungs-Nr.

EP4807476

Anmeldetag

6. März 2026

Veröffentlichung

16. September 2026

Rechtsraum

EP

IPC

G05B23/02

Abstract

A method (500) for conditional predictive maintenance in aircraft systems (100) using machine learning-based surrogate models (210) is disclosed. The described method (500) addresses the challenge of accurately detecting anomalies and predicting maintenance needs amidst complex operational and environmental variables. The method (500) involves receiving operational data from aircraft sensors (102), aircraft devices (104), and/or a memory (212), training a surrogate model (210) to predict aircraft device operations, and predicting operational outputs. Deviations (406) between predicted and actual outputs (400, 402) are analyzed to identify potential anomalies, distinguishing between noise and actual faults. This approach enables real-time monitoring and proactive maintenance scheduling, reducing reliance on scheduled inspections and enhancing aircraft reliability. The system (100) is adaptable to various environmental conditions and operational modes, ensuring consistent performance across different flight profiles. This method (500) is particularly useful for improving the efficiency and accuracy of aircraft maintenance operations.

Anmelder

Firma

The Boeing Company

Land

🇺🇸 USA

Fachgebiete