2021
DOI: 10.48550/arxiv.2112.03765
|View full text |Cite
Preprint
|
Sign up to set email alerts
|

In-flight Novelty Detection with Convolutional Neural Networks

Abstract: Gas turbine engines are complex machines that typically generate a vast amount of data, and require careful monitoring to allow for cost-effective preventative maintenance. In aerospace applications, returning all measured data to ground is prohibitively expensive, often causing useful, high value, data to be discarded. The ability to detect, prioritise, and return useful data in real-time is therefore vital. This paper proposes that system output measurements, described by a convolutional neural network model… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
4
1

Citation Types

0
7
0

Year Published

2022
2022
2023
2023

Publication Types

Select...
2

Relationship

0
2

Authors

Journals

citations
Cited by 2 publications
(7 citation statements)
references
References 25 publications
0
7
0
Order By: Relevance
“…Since the approximation of the aleatoric uncertainty according to [32] has already been successfully applied to in-flight measurements, it is used as starting point for further improvement. The approximation of the aleatoric uncertainty introduces additional model complexity to the neural network by requiring an increased number of output nodes.…”
Section: Artificial Neural Network With Uncertainty Quantificationmentioning
confidence: 99%
See 2 more Smart Citations
“…Since the approximation of the aleatoric uncertainty according to [32] has already been successfully applied to in-flight measurements, it is used as starting point for further improvement. The approximation of the aleatoric uncertainty introduces additional model complexity to the neural network by requiring an increased number of output nodes.…”
Section: Artificial Neural Network With Uncertainty Quantificationmentioning
confidence: 99%
“…The aleatoric uncertainty can be evaluated by approximating the probability density functions of individual measurements with artificial neural networks [31]. Despite an existing concept for approximating the aleatoric uncertainty for full-flight engine data [32], there is no method taking both the aleatoric and epistemic uncertainty into account.…”
Section: Introductionmentioning
confidence: 99%
See 1 more Smart Citation
“…Currently, most airlines adopted Quick Access Recorders (QAR) for data acquisition, providing flight data continuously sampled at frequencies of 1 Hz and more, which is also referred to as full-flight data covering the whole flight. The availability of these data obtained from a large variety of sensors enables the introduction of new methodologies to assess engine condition, which offers the chance to detect engine faults within one flight more reliably to support more efficient in-service operations and maintenance decisions [ 6 , 7 ]. An approach for fault detection based on steady-state flight regimes of full-flight data is demonstrated in [ 8 , 9 ].…”
Section: Introductionmentioning
confidence: 99%
“…Hartwell et al propose a practical and computationally inexpensive method for in-flight real-time anomaly detection based on a convolutional neural network. The efficacy of the method has been demonstrated on both real-time-series and synthetic snapshot data [ 7 ].…”
Section: Introductionmentioning
confidence: 99%