The 2nd International Conference On Cross-Disciplinary Academic Research 2023 , ICAR'2023 - 2023 : The Application of Machine Learning in Detecting Damage Structures

Farah Farzana, Abdul Aziz and Prof. Ts. Dr. Zahrah, Yahya and PM Dr Dewi Izzwi, Abdul Manan (2023) The 2nd International Conference On Cross-Disciplinary Academic Research 2023 , ICAR'2023 - 2023 : The Application of Machine Learning in Detecting Damage Structures. The 2nd International Conference On Cross-Disciplinary Academic Research 2023 , ICAR'2023, 2 (2). p. 92. ISSN 40181-700982

Text 18. The Application of Machine Learning in Detecting Damage Structures.pdf Download (3MB)

Abstract

The occurrence and development of damage to engineering structures is still not fully understood and is very difficult to be detected, located and quantified. Structural health monitoring (SHM) techniques using vibration-based damage detection (VBDD) are defined as a process to detect, locate and quantify structural damage. However, for a large, complex structure with joints, the VBDD method requires some knowledge of the damage location, which is itself a costly and time-consuming procedure due to the complex behaviour of the structural joints. This research aims to propose an advanced AI-based solution by using Machine Learning techniques to detect the presence of damage in the jointed structure. The newly developed algorithm of Machine Learning, which uses Artificial Neural Networks (ANNs) and modal testing techniques, is used to effectively and accurately predict the presence of damage in the structure. The Machine Learning algorithm is used to build the new prediction model and establish maximum correlation with the target response of the frequency response function (FRF) associated with the ensemble parameters. algorithm can learn from the data itself and find the optimal set for the given target. This leads to significant savings in terms of experimental effort, computational efficiency and faster decision making in detecting damage in the jointed structure. The result of the research, the development of the algorithm algorithms, will be able to learn from the data of the damaged structure itself and make predictions based on the learned data of the generated model

Item Type:	Article
Uncontrolled Keywords:	Machine Learning, Structural health monitoring, vibration-based damage detection, Artificial Neural Networks
Divisions:	Institute of Graduate Studies (IGS)
Depositing User:	LIBRARY2 UPTM
Date Deposited:	21 Jul 2025 01:52
Last Modified:	21 Jul 2025 01:52
URI:	http://eprints.kuptm.edu.my/id/eprint/4903

Actions (login required)

View Item