.png){kind=icon}
.png){kind=icon}
Return
Data-driven prediction of effective stiffness for thin-walled hollow piers via a GA-BP neural network with quasi-static test validation
.png){kind=icon}
PRE
AI
H
C
W
Y
Q
C
DOI：10.1016/j.advengsoft.2026.104194.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• GA-BP neural network integrates quasi-static tests for effective stiffness prediction. • Genetic algorithm optimizes BP network, reducing prediction error by 63 %. • Shear span ratio and axial compressive ratio dominate stiffness degradation (75.4 %). • Model validated with 4.58 % mean absolute error on experimental data. • Data-driven approach transforms seismic design from empirical to intelligent methods.
Keywords:
GA-BP neural network
effective stiffness prediction
quasi-static tests
genetic algorithm
shear span ratio
Journal
IF:
5.7
Papers:
3.3K
Citations:
1.2W
Organization
Cited Papers