August 15, 2025BatteriesOpen Access

State of Health Prediction of Lithium-Ion Batteries Based on Dual-Time-Scale Self-Supervised Learning

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Self-supervised learning estimates battery state of health using minimal labeled data, suggesting improved accuracy.

This method reduces labeling burden for battery health monitoring by 90–98%, enabling better industry applications.
On standard datasets, using only 10% labeled data, an RMSE of 0.491% was achieved for NCA estimation.
The dual-time-scale self-supervised learning framework enables automatic extraction of aging features.
Integrating domain knowledge into the attention mechanism enhances the accuracy of state of health estimation.

Li et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark