Accurate prediction of the State of Health (SOH) of lithium-ion batteries is essential for ensuring their safe and reliable operation. However, traditional deep learning approaches often suffer from challenges such as overfitting, limited generalization capability, and suboptimal prediction accuracy. To address these issues, this paper proposes a novel framework that combines a Long Short-Term Memory (LSTM) network with a Physics-Informed Neural Network (PINN), referred to as LSTM-PINN, for high-precision SOH estimation. The proposed framework models battery degradation using state-space equations and extracts latent temporal features. These features are further integrated into a Deep Hidden Temporal Physical Module (DeepHTPM), which incorporates physical prior knowledge into the learning process. This integration significantly enhances the model’s ability to accurately capture the complex dynamics of battery degradation. The effectiveness of LSTM-PINN is validated using two publicly available datasets based on graphite cathode materials (NASA and CACLE). Extensive experimental results demonstrate the superior predictive performance of the proposed model, achieving Mean Absolute Errors (MAEs) of just 0.594% and 0.746% and Root Mean Square Errors (RMSEs) of 0.791% and 0.897% on the respective datasets. Our proposed LSTM-PINN framework enables accurate battery aging modeling, advancing lithium-ion battery SOH prediction for industrial applications.

Accurately estimating the state of health (SOH) of lithium-ion batteries is vital for guaranteeing battery safety and prolonging their operational lifespan. However, current data-driven approaches often suffer from limited utilization of health indicators, weak noise suppression, and inefficient feature fusion across multiple branches. To overcome these limitations, this paper proposes DFFNet, a Dual-Branch Feature Fusion Network with Improved Dynamic Elastic Weight Consolidation (IDEWC) for battery SOH estimation. The proposed DFFNet captures both local degradation behaviors and global aging trends through two parallel branches, while a gated attention mechanism adaptively integrates multi-scale features. Moreover, IDEWC dynamically updates the Fisher Information Matrix to retain previously learned knowledge and adapt to new data, thereby mitigating catastrophic forgetting. Experimental validation on the NASA and CALCE datasets shows that DFFNet outperforms baseline methods in accuracy, demonstrating its robustness and generalizability for precise battery SOH estimation.