Aircraft Parametric Structural Load Monitoring Using Gaussian Process Regression

Abstract : The work presented here demonstrates the capability of Gaussian Process (GP) regression for the prediction of aircraft structural loads based on recorded flight parameters. The objective of monitoring aircraft loads during operation is to develop a better understanding of aircraft usage and thus to provide the operator with an accurate estimate of the remaining useful life of components against their prescribed fatigue life. These loads are often difficult and expensive to measure and this motivates the use of advanced mathematical techniques to estimate them accurately based on other parameters that are typically measured during flight. Gaussian Process regression is a powerful Bayesian machine learning tool whereby predictions and their distributions can be obtained without having to specify a particular model/functional form. Data collected from a military trainer aircraft is used to demonstrate how the mapping of measured strains to basic flight parameters such as airspeed, accelerations, and control surface deflections can be performed using GP regression. It is also shown how these results can be applied in an aircraft usage monitoring context. The GP predictions for strain are compared to the actual measurements for 101 flights, and the results are presented in terms of fatigue life, correlation, and mean-squared error. The results are encouraging, with errors in fatigue life in the range of 5% to 30% in the worst cases.
Type de document :
Communication dans un congrès
Le Cam, Vincent and Mevel, Laurent and Schoefs, Franck. EWSHM - 7th European Workshop on Structural Health Monitoring, Jul 2014, Nantes, France. 2014
Liste complète des métadonnées

Littérature citée [9 références]  Voir  Masquer  Télécharger

https://hal.inria.fr/hal-01022048
Contributeur : Anne Jaigu <>
Soumis le : jeudi 10 juillet 2014 - 09:59:10
Dernière modification le : mercredi 31 janvier 2018 - 15:14:02
Document(s) archivé(s) le : vendredi 10 octobre 2014 - 11:00:52

Fichier

0233.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01022048, version 1

Collections

Citation

Ramon Fuentes, Elizabeth Cross, Andrew Halfpenny, Keith Worden, Robert J. Barthorpe. Aircraft Parametric Structural Load Monitoring Using Gaussian Process Regression. Le Cam, Vincent and Mevel, Laurent and Schoefs, Franck. EWSHM - 7th European Workshop on Structural Health Monitoring, Jul 2014, Nantes, France. 2014. 〈hal-01022048〉

Partager

Métriques

Consultations de la notice

238

Téléchargements de fichiers

891