.png){kind=icon}
.png){kind=icon}
Return
Transformer-based prediction of free calcium oxide in cement clinker with channel–spatial dual-attention generative adversarial data augmentation
.png){kind=icon}
PRE
AI
Y
X
X
X
L
DOI：10.1016/j.jtice.2026.106871.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Dual-attention generative augmentation enriches sparse cement process data. • Transformer predicts free calcium oxide from 10 s process measurements. • RMSE and MAE for free calcium oxide are 0.0574 and 0.0391 wt%. • MAPE falls by 60.7% and coefficient of determination reaches 0.7513. • A 372,674-parameter model enables 1.0 ms per-sample inference.
Keywords:
Quality prediction
Data enhancement
Transformer
Generative adversarial networks
Channel–spatial dual-attention
Journal
IF:
6.3
Papers:
6.3K
Citations:
2.1W
Organization
No organization information available
Cited Papers