.png){kind=icon}
.png){kind=icon}
Return
Physics-Informed Unified Network for Robust and Interpretable SOH Estimation of Lithium-ion Batteries
.png){kind=icon}
PRE
AI
Y
Q
Y
R
W
W
J
DOI：10.1016/j.energy.2026.141447.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Propose a Physics-Informed Unified Network (PIUN) unifying data-driven learning with physical consistency for lithium-ion battery SOH estimation. • Design a unified gated backbone (UCBNet) that adaptively captures static health indicators and dynamic degradation behaviors. • Introduce physical-consistency regularization and PDE-based constraints to ensure realistic monotonic degradation while tolerating minor capacity regeneration phenomena. • Validate superior accuracy and cross-chemistry generalization across four public datasets (XJTU, TJU, MIT, HUST). • Reveal interpretable degradation mechanisms and seven key physical features through SHAP-based analysis.
Keywords:
Physics-Informed Unified Network
State-of-Health Estimation
Lithium-ion Batteries
Physical Consistency
Interpretable Models