Application of machine learning-base k-means clustering for feature recognition of fibre bragg grating acoustic signal from different delamination sizes
Continuous monitoring of composite structures is crucial to preserve their integrity over their entire service life, particularly when it comes to detecting subtle interior degradation like delamination. Extensive research has been dedicated to examining the utilisation of conventional electrical se...
| Main Authors: | , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English English |
| Published: |
IEEE
2024
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/42621/1/Application_of_Machine_Learning-Base_K-Means_Clustering_for_Feature_Recognition_of_Fibre_Bragg_Grating_Acoustic_Signal.pdf http://umpir.ump.edu.my/id/eprint/42621/2/Application_of_Machine_Learning-Base_K-Means_Clustering_for_Feature_Recognition_of_Fibre_Bragg_Grating_Acoustic_Signal_from_Different_Delamination_Sizes.pdf |
| Summary: | Continuous monitoring of composite structures is crucial to preserve their integrity over their entire service life, particularly when it comes to detecting subtle interior degradation like delamination. Extensive research has been dedicated to examining the utilisation of conventional electrical sensors for the purpose of collecting acoustic waves to quantify delamination. However, electrical sensors are well known to have several drawbacks. In this work, fibre Bragg grating (FBG) sensor is used to assess delamination, as the alternative to conventional electrical based sensors. In this study, composite plates were fabricated with varying sizes of delamination. The composite specimen has been equipped with a sensor network consisting of two multiplexed FBGs to acquire acoustic signals from an impact at the centre of the specimens. The use of a classification model derived from the analysis of sound signals from the FBGs has demonstrated significant success in identifying and characterising the result from different delamination sizes. The root-mean-square and peak value of the signals were extracted, and classification models were developed using these data sets. The results reveal that the overall percentage of accuracy is 91.7% for various delamination sizes. The findings offer compelling proof that employing an FBG sensor network for detecting delamination could be a practical choice for monitoring the health of plate-like composite structures. |
|---|