A Convolutional Neural Network Based Framework for Health Monitoring of a Welded Joint Steel Frame Structure

Maloth Naresh, Shirsendu Sikdar, Joy Pal

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Citations (Scopus)

Abstract

This study presents a Convolutional Neural Network (CNN) based Deep-learning framework for the health monitoring of a welded joint steel plane frame structure. The deep learning algorithm model is trained to extract local features from the vibration-based time-frequency scalogram images and using those features to distinguish the undamaged and two different damage cases of the steel plane frame structure. The performance and robustness of the framework are tested for an unseen image set corresponding to the training classes. The proposed deep learning framework can successfully classify the undamaged and damaged classes with high testing accuracy that signifies its efficiency as an automation tool for the health monitoring of steel plane frame structures subjected to damage near joints.

Original languageEnglish
Title of host publicationAdvances in Structural Mechanics and Applications
Subtitle of host publicationProceedings of ASMA-2021
EditorsJosé António Fonseca de Oliveira Correia, Satyabrata Choudhury, Subhrajit Dutta
PublisherSpringer, Cham
Pages251-262
Number of pages12
Volume26
Edition1st
ISBN (Electronic)9783030983352
ISBN (Print)9783030983345, 9783030983376
DOIs
Publication statusPublished - 9 Jun 2022
Externally publishedYes
EventInternational Conference on Advances in Structural Mechanics and Applications - Silchar, India
Duration: 26 Mar 202128 Mar 2021

Publication series

NameStructural Integrity
PublisherSpringer
Volume26
ISSN (Print)2522-560X
ISSN (Electronic)2522-5618

Conference

ConferenceInternational Conference on Advances in Structural Mechanics and Applications
Abbreviated titleASMA 2021
Country/TerritoryIndia
CitySilchar
Period26/03/2128/03/21

Cite this