Automated CFRP impact damage detection with statistical thermographic data and machine learning

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23640%2F25%3A43974591" target="_blank" >RIV/49777513:23640/25:43974591 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1016/j.ijthermalsci.2024.109411" target="_blank" >https://doi.org/10.1016/j.ijthermalsci.2024.109411</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.ijthermalsci.2024.109411" target="_blank" >10.1016/j.ijthermalsci.2024.109411</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Automated CFRP impact damage detection with statistical thermographic data and machine learning

Popis výsledku v původním jazyce

The study is focused on the use of machine learning models for the automated detection of impact damage in carbon fiber reinforced polymer (CFRP) by flash-pulse thermographic testing. A new method for thermographic data pre-processing, which is based on statistical features, was proposed. Nine machine learning models for the automated detection of impact damage in CFRP samples were applied to the raw thermographic data, data preprocessed by the suggested method and data pre-processed by the widely used thermographic signal reconstruction (TSR) method. The machine learning models were tested to provide a binary classification of impact damage in CFRP. The results presented in this study show improved performance of the classification if the data are pre-processed by the proposed method. The best results were obtained by a Bagged tree ensemble trained with statistical features. The final balanced accuracy achieved for the Bagged trees model trained on 40 statistical features was 99.8 % which indicates a very good performance.

Název v anglickém jazyce

Automated CFRP impact damage detection with statistical thermographic data and machine learning

Popis výsledku anglicky

The study is focused on the use of machine learning models for the automated detection of impact damage in carbon fiber reinforced polymer (CFRP) by flash-pulse thermographic testing. A new method for thermographic data pre-processing, which is based on statistical features, was proposed. Nine machine learning models for the automated detection of impact damage in CFRP samples were applied to the raw thermographic data, data preprocessed by the suggested method and data pre-processed by the widely used thermographic signal reconstruction (TSR) method. The machine learning models were tested to provide a binary classification of impact damage in CFRP. The results presented in this study show improved performance of the classification if the data are pre-processed by the proposed method. The best results were obtained by a Bagged tree ensemble trained with statistical features. The final balanced accuracy achieved for the Bagged trees model trained on 40 statistical features was 99.8 % which indicates a very good performance.

Druh

J_imp - Článek v periodiku v databázi Web of Science

CEP obor

—

OECD FORD obor

20501 - Materials engineering

Projekt

<a href="/cs/project/EF18_069%2F0010018" target="_blank" >EF18_069/0010018: LABIR-PAV / Předaplikační výzkum infračervených technologií</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)<br>I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2025

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název periodika

International Journal of Thermal Sciences

ISSN

1290-0729

e-ISSN

1778-4166

Svazek periodika

208

Číslo periodika v rámci svazku

FEB 2025

Stát vydavatele periodika

FR - Francouzská republika

Počet stran výsledku

21

Strana od-do

nestránkováno

Kód UT WoS článku

001317646200001

EID výsledku v databázi Scopus

2-s2.0-85203824289