.png){kind=icon}
.png){kind=icon}
Return
Gas-liquid two-phase flow pattern identification via DAS and temporal-enhanced ViT
.png){kind=icon}
PRE
AI
M
S
Y
Y
T
DOI：10.1016/j.ijmultiphaseflow.2026.105745.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• DAS technology combined with ViT achieves non-invasive flow pattern identification • ASLT transforms DAS signals into spatial-temporal-frequency 3D feature maps • Multi-frame temporal fusion improves accuracy from 95.97% to 98.42% • 5-frame model significantly enhances stratified flow classification by 4.68% • Ablation study quantifies temporal window impact across five flow patterns
Keywords:
DAS
ViT
flow pattern identification
temporal fusion
gas-liquid two-phase flow
Journal
IF:
3.8
Papers:
4.7K
Citations:
1.5W
Organization
Cited Papers