.png){kind=icon}
.png){kind=icon}
Return
Traffic state reconstruction in heterogeneous traffic flow from multi-source partial observations: A self-supervised and physics-informed framework
.png){kind=icon}
PRE
AI
J
D
W
L
B
DOI：10.1016/j.physa.2026.131785.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• A source-aware physics-enhanced grid fusion framework is proposed for mixed-traffic state reconstruction. • Multi-source CAV and RSU observations are unified through a time-space grid representation. • Block-wise self-supervision improves reconstruction under realistic continuous missing patterns. • Physics-guided constraints enhance congestion propagation consistency in the time–space plane.
Journal
P
IF:
3.1
Papers:
1.3K
Citations:
3.6W
Organization
Cited Papers