Search

Lithium-ion batteries are widely utilized in numerous applications, making it essential to precisely predict their degradation trajectory and remaining useful life (RUL). To improve the stability and applicability of RUL prediction for lithium-ion batteries, this paper uses a new method to predict RUL by combining CNN-LSTM-Attention with transfer learning. The presented model merges the strengths of both convolutional and sequential architectures, and it enhances the model’s capability to grasp comprehensive information by utilizing the attention mechanism, thereby boosting overall performance. The CEEMDAN algorithm is used for NASA batteries with obvious capacity regeneration phenomena to alleviate the difficulties caused by capacity regeneration on model prediction. During the model transfer phase, the CNN and LSTM layers of the pre-trained model from the source domain are kept unchanged during retraining, while the attention and fully connected layers are fine-tuned for NASA batteries and self-tested NCM batteries. The final results indicate that this method achieves superior accuracy relative to other methods while addressing the issue of limited labeled data in the target domain through transfer learning, thereby enhancing the model’s transferability and generalization capabilities.

The electrochemical impedance spectrum (EIS) is a non-destructive technique for the on-line evaluation and monitoring of the performance of lithium-ion batteries. However, the measured EIS can be unstable and inaccurate without the proper resting time. Therefore, we conducted comprehensive EIS tests during the charging process and at different state of charge (SOC) levels with various resting times. The test results revealed two findings: (1) EIS tests with a constant long resting time showed a clear pattern in the impedance spectral radius—a decrease followed by a slight increase. We analyzed the impedance data using an equivalent circuit model and explained the changes through circuit parameters. (2) We examined the effect of resting time on impedance at consistent SOC levels. While low SOC levels exhibited significant sensitivity to resting time, medium SOC levels showed less sensitivity, and high SOC levels had minimal impact on resting time. The equivalent circuit parameters matched the observed trends. Kramers–Kronig transformation was conducted to assess the reliability of the experiments. This study not only summarizes the relationship between the EIS and SOC but also highlights the importance of resting time in impedance analysis. Recognizing the role of the resting time could enhance impedance-based battery studies, contribute to refined battery status evaluation, and help researchers to design proper test protocols.