Developing robust stress intensity factor models using Fourier-based data analysis to guide machine learning method selection and training

doi:10.1016/j.engfracmech.2025.111387

Developing robust stress intensity factor models using Fourier-based data analysis to guide machine learning method selection and training

2025-07-19

0

PRE

AI

DOI：10.1016/j.engfracmech.2025.111387

Original

Original request for help

Save

Abstract

En 中文

• Machine learning (ML) offers an alternative way to create SIF models. • The first application of deep operator network (DeepONet) to the prediction of SIFs. • DeepONet improves the accuracy compared to other commonly used ML models. • Guidelines for ML model selection based on dataset analysis using Fourier transform (FT). • FT helps guide ML model selection by analyzing frequency decay patterns.

Keywords:

machine learning

stress intensity factor

DeepONet

Fourier transform

model selection

Developing robust stress intensity factor models using Fourier-based data analysis to guide machine learning method selection and training

Abstract

Journal

Organization