Variational Neural Network Embedded with Digital Twins for Probabilistic Structural Damage Quantification

AIAA Journal, Год журнала: 2024, Номер unknown, С. 1 - 13

Опубликована: Дек. 17, 2024

Quantifying structural damage using online monitoring data is crucial for condition-based maintenance to ensure aviation safety. However, most data-driven methods hardly use accumulated domain knowledge, making it difficult address parameter variability across different structures due manufacturing as well compromising result interpretability. To these challenges, this study proposes a physics-decoded variational neural network quantification and model calibration. The innovation of method lies in seamlessly integrating reduced-order digital twin containing states influencing parameters decoder within the training physical feature extraction inference. This architecture enables individualized, real-time calibration an entire fleet, while accounting uncertainties. Validation on typical damaged aeronautical panels demonstrates that proposed accurately predicts quantifies associated uncertainties, thereby ensuring high interpretability accuracy. approach expected be integrated into airframe framework enable fleet.

Язык: Английский

A general probabilistic framework for impact localisation based on flexural wave propagation

Mechanical Systems and Signal Processing, Год журнала: 2025, Номер 226, С. 112320 - 112320

Опубликована: Янв. 16, 2025

Язык: Английский

Data fusion based impact localization on aircraft composite structures using Bayesian estimation and weighted averaging

Mechanical Systems and Signal Processing, Год журнала: 2025, Номер 230, С. 112675 - 112675

Опубликована: Апрель 6, 2025

Язык: Английский