Damage Detection in Rods via Use of a Genetic Algorithm and a Deep-Learning Based Surrogate
METADATA ONLY
Loading...
Author / Producer
Date
2023
Publication Type
Conference Paper
ETH Bibliography
yes
Citations
Altmetric
METADATA ONLY
Data
Rights / License
Abstract
This research demonstrates the use of genetic algorithms for damage detection in isotropic rods. The spectral element method and a deep-learning-based surrogate model is utilized for simulating wave propagation in an isotropic cracked rod. The genetic algorithm employs results (“numerical experiment") obtained from the spectral element model and the deep-learning-based surrogate to determine the optimized crack locations and crack depths as output parameters. The objective function used in the genetic algorithm is the mean square error between the response obtained from spectral element model and the deep-learning-based surrogate model.
Permanent link
Publication status
published
External links
Book title
European Workshop on Structural Health Monitoring
Journal / series
Volume
270
Pages / Article No.
272 - 280
Publisher
Springer
Event
10th European Workshop on Structural Health Monitoring (EWSHM 2022)
Edition / version
Methods
Software
Geographic location
Date collected
Date created
Subject
Wave propagation; Deep-learning; Damage detection; Spectral element method; Convolution neural networkds (CNN); Genetic algorithm
Organisational unit
03890 - Chatzi, Eleni / Chatzi, Eleni