English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Compared with traditional model-based fault detection and classification(FDC) methods, deep neural networks (DNN) prove to be effective for theaerospace sensors FDC problems. However, time being consumed in training theDNN is excessive, and explainability analysis for the FDC neural network isstill underwhelming. A concept known as imagefication-based intelligent FDC hasbeen studied in recent years. This concept advocates to stack the sensorsmeasurement data into an image format, the sensors FDC issue is thentransformed to abnormal regions detection problem on the stacked image, whichmay well borrow the recent advances in the machine vision vision realm.Although promising results have been claimed in the imagefication-basedintelligent FDC researches, due to the low size of the stacked image, smallconvolutional kernels and shallow DNN layers were used, which hinders the FDCperformance. In this paper, we first propose a data augmentation method whichinflates the stacked image to a larger size (correspondent to the VGG16 netdeveloped in the machine vision realm). The FDC neural network is then trainedvia fine-tuning the VGG16 directly. To truncate and compress the FDC net size(hence its running time), we perform model pruning on the fine-tuned net. Classactivation mapping (CAM) method is also adopted for explainability analysis ofthe FDC net to verify its internal operations. Via data augmentation,fine-tuning from VGG16, and model pruning, the FDC net developed in this paperclaims an FDC accuracy 98.90% across 4 aircraft at 5 flight conditions (runningtime 26 ms). The CAM results also verify the FDC net w.r.t. its internaloperations.

Fault Detection and Classification of Aerospace Sensors using a VGG16-based Deep Neural Network