The estimation of the state of health (SOH) of lithium-ion batteries is a critical technology for enhancing battery lifespan and safety. When estimating SOH, it is essential to select representative features, commonly referred to as health indicators (HIs). Most existing studies primarily focus on HIs related to capacity degradation and internal resistance increase. However, due to the complexity of lithium-ion battery degradation mechanisms, the relationships between these mechanisms and health indicators remain insufficiently explored. This paper provides a comprehensive review of core methodologies for SOH estimation, with a particular emphasis on the classification and extraction of health indicators, direct measurement techniques, model-based and data-driven SOH estimation approaches, and emerging trends in battery management system applications. The findings indicate that capacity, internal resistance, and temperature-related indicators significantly impact SOH estimation accuracy, while machine learning models demonstrate advantages in multi-source data fusion. Future research should further explore composite health indicators and aging mechanisms of novel battery materials, and improve the interpretability of predictive models. This study offers theoretical support for the intelligent management and lifespan optimization of lithium-ion batteries.

To address the challenge of simultaneously modelling temporal evolution and static properties in fatigue life prediction, this paper proposes a Hybrid GRU–Attention–DNN model: The Gated Recurrent Unit (GRU) captures time-evolution features, while the attention mechanism adaptively focuses on critical stages. These are then fused with static properties via a fully connected network to generate life estimates. Training and validation were conducted using an 8:2 split, with baselines including Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and GRU. Performance was evaluated using the coefficient of determination (R²), root mean squared error (RMSE), mean absolute error (MAE), and root mean squared logarithmic error (RMSLE), together with error band plots. Results demonstrate that the proposed model outperforms baseline CNN/GRU/LSTM models in overall accuracy and robustness, and that these improvements remain statistically significant according to bootstrap confidence intervals (CI) of R², RMSE, MAE and RMSLE on the test set. Additionally, this paper conducts an interpretability analysis: attention visualisations reveal the model’s significant emphasis on the early stages of the lifespan. Time window masking experiments further indicate that removing early information causes the most significant performance degradation. Both lines of evidence show high consistency in qualitative and quantitative trends, providing a basis for engineering sampling window design and trade-offs in test duration.