A Hybrid Model Based on Battery IC Curve and Combined Optimization Algorithm for Joint Prediction of SOH and RUL of Lithium‐Ion Batteries

Accurate state of health (SOH) and remaining useful life (RUL) prediction is thus the core function of reliable battery management systems. However, recursive neural networks face problems such as insufficient capture of key degraded features and inaccurate identification of hyperparameters, which lead to poor generalization ability. This paper proposes a hybrid RIME‐CNN‐BiLSTM‐Attention framework for joint SOH and RUL prediction. First, health indicators are extracted from the incremental capacity curves of battery charge–discharge cycles, and the Pearson correlation analysis is used to screen capacity‐sensitive features. Second, a novel hybrid prediction framework based on the RIME‐CNN‐BiLSTM‐Attention model is proposed to enhance focus on key features, and the RIME optimization algorithm is applied for automatic hyperparameter tuning. Finally, the ablation studies and multi‐dataset validation demonstrate the superior performance of the proposed method: the maximum prediction error is within 2.14% for SOH and 1.67% for RUL, with the minimum mean absolute error (MAE), mean absolute percentage error (MAPE), and root mean square error (RMSE) reaching 1.11%, 0.85%, 1.50% for SOH, and 0.91%, 1.20%, 1.22% for RUL, respectively, which demonstrate that the proposed approach achieves better accuracy and robustness in predicting SOH and RUL.