A Transferable State-of-Health Prediction Method for Lithium-ion Batteries Based on ConvLSTM-Trans

The accurate prediction of the state-of-health (SOH) of lithium-ion batteries(LIBs) is crucial for ensuring their reliability and longevity in various applications, such as electric vehicles and energy storage systems. In this paper, we propose a novel transferable ConvLSTM-Trans model to address the challenges in predicting the SOH of LIBs. Our approach combines Convolutional Neural Networks (CNN) for feature extraction, Long Short-Term Memory (LSTM) networks for capturing temporal dependencies, and Transformer architecture for global sequence modeling, enabling it to effectively handle the complex degradation patterns of battery data. We evaluate the model on two publicly available datasets, CALCE and NASA, and show that it outperforms both standalone and hybrid models in terms of RMSE, MAE, and R2 scores. Additionally, we demonstrate the effectiveness of transfer learning with the Maximum Mean Discrepancy (MMD) loss in enhancing the model’s generalization ability. Hyperparameter tuning further improves performance, highlighting the importance of optimization for achieving accurate SOH predictions. Our results show that the proposed model provides a promising solution for enhancing the accuracy of lithium-ion battery health prediction.