A Multi-Encoder BHTP Autoencoder for Robust Lithium Battery SOH Prediction Under Small-Sample Scenarios
Abstract
:1. Introduction
- (1)
- We propose a novel multi-encoder framework for SOH prediction of lithium-ion batteries, which integrates element, positional, and temporal encoders to extract multidimensional features, enabling comprehensive and accurate analysis of SOH sequences under small-sample conditions.
- (2)
- We develop an innovative BHTP module for feature fusion and encoding optimization, incorporating residual connections, convolutional layers, and layer normalization, which significantly enhances the model’s representational capability, mitigates overfitting risks, and improves stability and generalization.
- (3)
- We propose a transfer learning strategy that combines pretraining with fine-tuning, optimizing model parameters for small-sample scenarios, significantly improving prediction accuracy while reducing computational complexity.
- (4)
- The experiments on NASA datasets show that the proposed method excels under small-sample conditions, achieving fast convergence, significant reductions in RMSE and MAE, and accurate SOH predictions, demonstrating its effectiveness and practicality.
2. Related Works
3. Multi-Encoder Architecture and BHTP-Enhanced Feature Fusion
3.1. Task Definitions
3.2. Sliding Window Prediction
3.3. Multi-Encoder Architecture: Overall Structural Design
3.4. Advanced Multi-Encoder Architecture
3.5. Residual-Fused BHTP Module
3.6. Adaptive and Hybrid Transfer Learning
4. Experiment and Results Analysis
4.1. Dataset and Experimental Setup
4.2. Evaluation Indicators
4.3. Convergence Experiment
4.4. Comparative Experiment
4.5. Ablation Experiment
4.6. Case Study
5. Conclusions
Author Contributions
Funding
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- Hong, J.; Li, K.; Liang, F.; Yang, H.; Zhang, C.; Yang, Q.; Wang, J. A novel state of health prediction method for battery system in real-world vehicles based on gated recurrent unit neural networks. Energy 2024, 289, 129918. [Google Scholar] [CrossRef]
- Wei, Y.; Wu, D. State of health and remaining useful life prediction of lithium-ion batteries with conditional graph convolutional network. Expert Syst. Appl. 2024, 238, 122041. [Google Scholar] [CrossRef]
- Zhang, X.Y.; Wu, J.H.; Zhang, L.P.; Dai, H.D.; Bai, W.Q.; Lin, H.J.; Zhang, F.; Yang, Y.X. Research progress of lithiumion battery health state assessment method. Henan Sci. 2024, 42, 1717–1740. [Google Scholar]
- Buchanan, S.; Crawford, C. Probabilistic lithium-ion battery state-of-health prediction using convolutional neural networks and Gaussian process regression. J. Energy Storage 2024, 76, 109799. [Google Scholar] [CrossRef]
- Wang, X.; Hu, B.; Su, X.; Xu, L.; Zhu, D. State of health estimation for lithium-ion batteries using random forest and gated recurrent unit. J. Energy Storage 2024, 76, 109796. [Google Scholar] [CrossRef]
- Che, Y.; Zheng, Y.; Forest, F.E.; Sui, X.; Hu, X.; Teodorescu, R. Predictive health assessment for lithium-ion batteries with probabilistic degradation prediction and accelerating aging detection. Reliab. Eng. Syst. Saf. 2024, 241, 109603. [Google Scholar] [CrossRef]
- Wu, T.Z.; Zhu, J.C.; Cheng, X.F.; Kang, J. SOH estimation method for lithium battery Based on charging phase data and GWO-BiLSTM model. Power Supply Technol. 2024, 48, 2184–2194. [Google Scholar]
- Liu, P.; Li, Z.W.; Cai, Y.S.; Wang, W.; Xia, X.Y. Based on the equivalent circuit model and data driven fusion model of SOC and SOH joint estimation approach. J. Electrotech. 2024, 39, 3232–3243. [Google Scholar]
- Wu, L.X.; Pang, H.; Jin, J.M.; Geng, Y.F.; Liu, K. Review of state-of-charge estimation methods for lithium-ion batteries based on electrochemical models. Trans. China Electrotech. Soc. 2022, 37, 1703–1725. [Google Scholar]
- Jin, J.X.; Yu, R.X.; Liu, G.; Xu, L.B.; Ma, Y.Q.; Wang, H.B.; Hu, C. Research progress on estimation methods for the state of health of lithium-ion batteries. J. Eng. Eng. 2024, 19, 33–48. [Google Scholar]
- Messing, M.; Shoa, T.; Habibi, S. Estimating battery state of health using electrochemical impedance spectroscopy and the relaxation effect. J. Energy Storage 2021, 43, 103210. [Google Scholar] [CrossRef]
- Shi, H.S.; Sun, X.W.; Wang, K. Health state estimation of lithium-ion battery based on electrochemical impedance spectroscopy. Power Gener. Technol. 2024, 1–15. Available online: http://kns.cnki.net/kcms/detail/33.1405.tk.20240716.1502.008.html (accessed on 8 March 2025).
- Luo, F.; Huang, H.; Ni, L.; Li, T. Rapid prediction of the state of health of retired power batteries based on electrochemical impedance spectroscopy. Chin. J. Sci. Instrum. 2021, 42, 172–180. [Google Scholar]
- Wang, Y.; Wei, Q.G.; Sun, P.; Zhang, X.F.; Liu, Y.J. Joint estimation of lithium-battery SOC and SOH based on dual extended kalman filter. In Proceedings of the 2022 China Society of Automotive Engineers Annual Conference, Nantong, China, 20–22 December 2022; pp. 4–9. [Google Scholar]
- Yang, D.; Zhang, X.; Pan, R.; Wang, Y.; Chen, Z. A novel gaussian process regression model for state-of-health estimation of lithium-ion battery using charging curve. J. Power Sources 2018, 384, 387–395. [Google Scholar] [CrossRef]
- Hui, Z.L.; Wang, R.J.; Feng, N.N.; Yang, M. State of health prediction of lithium-ion batteries based on ensemble gaussian process regression. J. Meas. Sci. Instrum. 2024, 15, 397–407. [Google Scholar] [CrossRef]
- Tan, Y.; Zhao, G. Transfer learning with long short-term memory network for state-of-health pre-diction of lithium-ion batteries. IEEE Trans. Ind. Electron. 2019, 67, 8723–8731. [Google Scholar] [CrossRef]
- Li, K.; Hu, L.; Song, T.T. State of health estimation of lithium-ion batteries based on CNN-Bi-LSTM. Shandong Electr. Power Technol. 2023, 50, 66–72. [Google Scholar]
- Ke, H. SOH estimation of lithium battery based on diffusion model and bidirectional long short-term memory network. Ploidy Henan Sci. Technol. 2024, 51, 5–11. [Google Scholar]
- Zheng, W.B.; Zhou, X.Y.; Wu, Y.; Feng, L.; Yin, H.T.; Fu, P. Experimental design for state of health (SOH) estimation of lithium batteries based on transfer learning. Exp. Technol. Manag. 2022, 39. [Google Scholar] [CrossRef]
- Yin, J.; Liu, B.; Sun, G.B.; Qian, X.W. Remaining useful life prediction of lithium-ion batteries based on transfer learning and denoising autoencoder-long short-term memory. Trans. China Electrotech. Soc. 2024, 39, 289–302. [Google Scholar]
- Mo, Y.M.; Yu, Z.H.; Ye, P.; Fan, W.J.; Lin, Y. SOH estimation of lithium battery based on transfer learning and GRU neural network. Acta Sol. Energy 2024, 45, 233–239. [Google Scholar]
- Cheng, G.; Wang, X.; He, Y. Remaining useful life and state of health prediction for lithium batteries based on empirical mode decomposition and a long and short memory neural network. Energy 2021, 232, 121022. [Google Scholar] [CrossRef]
- Zhang, X.F.; Chen, Y.L.; Li, S.Q.; Zeng, X.K.; Lian, X.; Huang, C. LSTM neural network based state estimation of lithium ion battery health. Auto Pract. Technol. 2025, 50, 1–6. [Google Scholar] [CrossRef]
- Chang, C.; Wang, Q.; Jiang, J.; Wu, T. Lithium-ion battery state of health estimation using the incremental capacity and wavelet neural networks with genetic algorithm. J. Energy Storage 2021, 38, 102570. [Google Scholar] [CrossRef]
- Huang, J.Y.; Bai, J.Q.; Xiang, Y.H. SOH estimation method for lithium-ion battery based on DOD-LN-GPR model. Chin. J. Sol. Energy 2025, 46, 60–69. [Google Scholar]
- Tian, A.N.; Yang, C.; Gao, Y.; Li, T.Y.; Wang, L.J.; Chang, C.; Jiang, J.C. A state of health estimation method of lithium-ion batteries based on DT-IC-V health features extracted from partial charging segment. Int. J. Green Energy 2023, 20, 997–1011. [Google Scholar] [CrossRef]
- Li, Z.L.; Qiao, G.Z.; Cui, F.S.; Cai, J.H.; Shi, Y.H.; Wang, B.H. Prediction of state of health of lithium-ion batteries using the VMD-HPO-NBEATS model. China Test 2024, 50, 65–73. [Google Scholar]
- Shi, S.S.; Gao, Z.B. Based on the automatic feature extraction and IWOA—Lithium-ion battery SOH prediction of SVR model. Intern. Combust. Engines Accessories 2024, 22, 10–13. [Google Scholar]
- Li, J.; Chen, X.L.; Xu, L. Li-ion battery health estimation based on multiple health factors and IPSO-LSTM model. Automot. Engine 2025, 1, 39–46. [Google Scholar]
- Zhang, M.; Yang, D.F.; Du, J.X.; Sun, H.L.; Li, L.W.; Wang, L.C.; Wang, K. A Review of SOH prediction of li-ion batteries based on data-driven algorithms. Energies 2023, 16, 3167. [Google Scholar] [CrossRef]
- Dong, X.H.; Dong, J.B.; Wang, M.S.; Zeng, F.; Pan, Y. Rapid estimation method of lithium battery health state Based on new health characteristics. Electr. Power Eng. Technol. 2025, 44, 136–142+206. [Google Scholar]
- Shi, Y.H.; Tian, J.Y.; Liu, J.J.; Yang, S.Q. Research on multi-source adaptive transfer learning algorithm based on manifold structure and its application. Control. Decis. 2023, 38, 797–804. [Google Scholar]
- Li, L.; Yan, X.M.; Zhang, Y.S.; Feng, Y.L.; Hu, H.L.; Duan, Y.J.; Cui, C.H. Lithium battery state of charge estimation method based on transfer learning. J. Xi’an Jiao Tong University 2023, 57, 142–150. [Google Scholar]
- Liu, B.; Yang, J.; Wang, R.G.; Xue, L.X. Memory-based transfer learning for few-shot learning. Comput. Eng. Appl. 2022, 58, 242–249. [Google Scholar]
- Chang, W.; Hu, Z.C.; Pan, D.Z.; Shi, J.W. Application of an improved VAE-GAN model to battery EIS data augmentation. Sci. Technol. Ind. 2024, 24, 258–263. [Google Scholar]
- Zhao, Y.P.; Huang, W.; Zhang, J.F. Lithium battery life prediction based on LSTM-Transformer with hybrid scale health factor. Electron. Meas. Technol. 2024, 11, 112–122. [Google Scholar]
- Li, D.C.; Wu, C.S.; Tsai, T.I.; Lina, Y.S. Using mega-trend-diffusion and artificial samples in small data set learning for early flexible manufacturing system scheduling knowledge. Comput. Oper. Res. 2007, 34, 966–982. [Google Scholar] [CrossRef]
- Qiao, J.F.; Guo, Z.H.; Tang, J. Virtual sample generation method based on improved metertrend diffusion and hidden layer interpolation and its application. Chin. Chem. Eng. J. 2020, 71, 5681–5695. [Google Scholar]
- Kang, G.; Wu, L.; Guan, Y.; Peng, Z. A Virtual sample generation method based on differential evolution Al-gorithm for overall trend of small sample data: Used for lithium-ion battery capacity degradation data. IEEE Access 2019, 7, 123255–123267. [Google Scholar] [CrossRef]
- Zhu, H.J.; Lü, Z.G.; Di, R.H.; Sun, X.J.; Hao, K.Q. To improve MD-MTD lithium battery life prediction neural network simulation. J. Xi’an Univ. Technol. 2022, 42, 620–626. [Google Scholar]
- Xu, Z.J.; Zhu, Q.; Xu, S.F.; Mao, Q. Research on parameter identification algorithm of lithium-battery model based on differential evolution method. Electr. Eng. 2021, 7, 35–37. [Google Scholar]
- Zhu, Y.F.; Jiang, G.P.; Gao, H.; Li, W.Z.; Gui, Y.C. Prediction of battery state of charge based on feature selection and data augmentation. Comput. Syst. Appl. 2023, 32, 45–54. [Google Scholar]
- Cui, X.K.; Wang, Q.Z.; Liu, Q.P. State of health prediction of lithium batteries based on data-driven approaches. Complex Syst. Complex. Sci. 2024, 21, 154–159. [Google Scholar]
- Liu, X.H.; Gao, Z.C.; Tian, J.Q.; Wei, Z.B.; Fang, C.Q.; Wang, P. State of health estimation for lithium-ion batteries using voltage curves reconstruction by conditional generative adversarial network. IEEE Trans. Transp. Electrif. 2024, 10, 10557–10567. [Google Scholar] [CrossRef]
- Ren, Y.; Tang, T.; Jiang, F.S.; Xia, Q.; Zhu, X.Y.; Sun, B.; Yang, D.Z.; Feng, Q.; Qian, C. A novel state of health estimation method for lithium-ion battery pack based on cross generative adversarial networks. Appl. Energy 2025, 377, 124385. [Google Scholar] [CrossRef]
- Guo, F.; Huang, G.S.; Zhang, W.C.; Wen, A.; Li, T.T.; He, H.C.; Huang, H.L.; Zhu, S.S. Lithium battery state-of-health estimation based on sample data generation and temporal convolutional neural network. Energies 2023, 16, 8010. [Google Scholar] [CrossRef]
- Lin, C.H.; Kaushik, C.; Dyer, E.L.; Muthukumar, V. The good, the bad and the ugly sides of data augmentation: An implicit spectral regularization perspective. J. Mach. Learn. Res. 2024, 25, 1–85. [Google Scholar]
- Wang, Q.H.; Jia, H.J.; Huang, L.X.; Mao, Q.R. Semantic contrastive clustering with federated data augmentation. J. Comput. Res. Dev. 2024, 61, 1511–1524. [Google Scholar]
- Thompson, M.L.; Kramer, M.A. Modeling chemical processes using prior knowledge and neural networks. Alche J. 1994, 40, 1328–1340. [Google Scholar] [CrossRef]
- Zhao, W.; Yan, H.; Zhou, Z. Transformer fault diagnosis based on a residual BP neural network. Electr. Power Autom. Equip. 2020, 310, 143–148. [Google Scholar]
- Hu, X.Q.; Geng, L.M.; Shu, J.H.; Zhang, W.B.; Wu, C.L.; Wei, X.L.; Huang, D.; Chen, H. Health state estimation of lithium-ion battery using global health factor and residual Model. J. Xi’an Jiaotong Univ. 2025, 59, 105–117. [Google Scholar]
- Hao, K.Q.; Lü, Z.G.; Li, Y.; Di, R.H.; Zhu, H.J. Remaining useful life prediction of lithium-ion batteries using BP neural network incorporating prior knowledge. J. Xi’an Technol. Univ. 2022, 42, 65–73. [Google Scholar]
- Fan, J.H.; Liu, Q.Y.; Ma, L.; Liu, L.H. Life prediction model of lithium-ion batteries for electric vehicles based on deep learning. J. Terahertz Sci. Electron. Inf. Technol. 2025, 23, 182–187. [Google Scholar]
- Jiang, B.; Zhu, J.G.; Wang, X.Y.; Wei, X.Z.; Shang, W.L.; Dai, H.F. A comparative study of different features extracted from electrochemical impedance spectroscopy in state of health estimation for lithium-ion batteries. Appl. Energy 2022, 322, 119502. [Google Scholar] [CrossRef]
- Li, J.H.; Li, T.T.; Qiao, Y.J.; Tan, Z.J.; Qiu, X.H.; Deng, H.; Li, W.; Qi, X.; Wu, W.X. Internal temperature estimation method for lithium-ion battery based on multi-frequency imaginary part impedance and GPR model. J. Energy Storage 2025, 118, 116287. [Google Scholar] [CrossRef]
- Li, Z.H.; Shi, Q.L.; Wang, K.L.; Jiang, K. Research status and prospect of lithium-ion battery health state estimation methods. Autom. Electr. Power Syst. 2024, 48, 109–129. [Google Scholar]
- Zhou, Y.F.; Sun, X.X.; Huang, L.J.; Lian, J. Life cycle oriented health state estimation of lithium battery. J. Harbin Inst. Technol. 2021, 53, 55–62. [Google Scholar]
- Zhang, S.F.; Zhang, Q.Y.; Yang, Y.S.; Su, Y.X.; Xiong, B.Y. Modeling method of lithium-ion battery based on sliding window and LSTM neural network. Energy Storage Sci. Technol. 2022, 11, 228–239. [Google Scholar]
- Chen, M.H.; Wang, T.; Yuan, Y.; Ke, S.T. Study on retinal OCT segmentation with dual-encoder. Opto-Electron. Eng. 2023, 50, 230146. [Google Scholar]
- Ju, F.J.; Wu, Y.C.; Dong, M.J.; Zhao, J.X. MiFDeU: Multi-information fusion network based on dual-encoder for pelvic bones segmentation. Eng. Appl. Artif. Intell. 2025, 147, 110230. [Google Scholar] [CrossRef]
- NASA Ames Prognostics Data Repository. Available online: https://www.nasa.gov/intelligent-systems-division/discovery-and-systems-health/pcoe/pcoe-data-set-repository/ (accessed on 10 April 2025).
- Van Houdt, G.; Mosquera, C.; Nápoles, G. A review on the long short-term memory model. Artif. Intell. Rev. 2020, 53, 5929–5955. [Google Scholar] [CrossRef]
- Huang, Z.; Xu, W.; Yu, K. Bidirectional LSTM-CRF models for sequence tagging. arXiv 2015, arXiv:1508.01991. [Google Scholar]
- Zargar, S. Introduction to Sequence Learning Models: RNN, LSTM, GRU. 2021. Available online: https://www.researchgate.net/profile/Sakib-Zargar-2/publication/350950396_Introduction_to_Sequence_Learning_Models_RNN_LSTM_GRU/links/607b41c0907dcf667ba83ade/Introduction-to-Sequence-Learning-Models-RNN-LSTM-GRU.pdf (accessed on 8 March 2025).
Dataset | Temperature | Condition of Charge | Condition of Release | Rated Capacity | End Condition |
---|---|---|---|---|---|
B0005 | 24 °C | 1.5 A:4.2 V → 20 mA | 2 A → 2.7 V | 2 Ahr | 1.4 Ahr |
B0007 | 24 °C | 1.5 A:4.2 V → 20 mA | 2 A → 2.2 V | 2 Ahr | 1.4 Ahr |
B0033 | 24 °C | 1.5 A:4.2 V → 20 mA | 4 A → 2.0 V | 2 Ahr | 1.6 Ahr |
Dataset | Model | Training Set Division | RMSE | MAE |
---|---|---|---|---|
B0007 | Ours | Top20% | 0.017 | 0.015 |
Top30% | 0.010 | 0.009 | ||
Top40% | 0.002 | 0.002 | ||
LSTM | Top20% | 0.057 | 0.051 | |
Top30% | 0.020 | 0.019 | ||
Top40% | 0.013 | 0.012 | ||
BiLSTM | Top20% | 0.056 | 0.051 | |
Top30% | 0.025 | 0.022 | ||
Top40% | 0.013 | 0.012 | ||
BiGRU | Top20% | 0.059 | 0.053 | |
Top30% | 0.026 | 0.023 | ||
Top40% | 0.011 | 0.010 | ||
GRU | Top20% | 0.038 | 0.034 | |
Top30% | 0.013 | 0.012 | ||
Top40% | 0.006 | 0.005 | ||
B0033 | Ours | Top20% | 0.007 | 0.007 |
Top30% | 0.004 | 0.005 | ||
Top40% | 0.002 | 0.003 | ||
LSTM | Top20% | 0.007 | 0.007 | |
Top30% | 0.011 | 0.010 | ||
Top40% | 0.005 | 0.005 | ||
BiLSTM | Top20% | 0.012 | 0.011 | |
Top30% | 0.007 | 0.007 | ||
Top40% | 0.008 | 0.008 | ||
BiGRU | Top20% | 0.015 | 0.013 | |
Top30% | 0.010 | 0.010 | ||
Top40% | 0.004 | 0.003 | ||
GRU | Top20% | 0.013 | 0.012 | |
Top30% | 0.007 | 0.007 | ||
Top40% | 0.004 | 0.003 |
Dataset | Temporal Encoder | Positional Encoder | Token Encoder | BHTP | RMSE | MAE |
---|---|---|---|---|---|---|
B0007 | √ | √ | √ | 0.012 | 0.010 | |
√ | √ | 0.013 | 0.010 | |||
√ | √ | √ | 0.015 | 0.014 | ||
√ | √ | √ | 0.017 | 0.015 | ||
B0033 | √ | √ | √ | 0.005 | 0.004 | |
√ | √ | 0.006 | 0.005 | |||
√ | √ | √ | 0.006 | 0.006 | ||
√ | √ | √ | 0.007 | 0.007 |
Dataset | Number of Cycles | Predicted Value | True Value | Difference |
---|---|---|---|---|
B0007 | 43 | 0.906 | 0.905 | +0.001 |
85 | 0.814 | 0.800 | +0.014 | |
127 | 0.767 | 0.746 | +0.021 | |
168 | 0.740 | 0.716 | +0.024 | |
B0033 | 35 | 0.802 | 0.798 | +0.004 |
89 | 0.732 | 0.724 | +0.008 | |
138 | 0.672 | 0.663 | +0.009 | |
197 | 0.667 | 0.658 | +0.009 |
Dataset | Number of Cycles | Predicted Value | True Value | Difference |
---|---|---|---|---|
B0007 | 58 | 0.872 | 0.873 | −0.001 |
95 | 0.803 | 0.795 | +0.008 | |
132 | 0.751 | 0.738 | +0.013 | |
168 | 0.730 | 0.716 | +0.014 | |
B0033 | 52 | 0.806 | 0.809 | −0.003 |
100 | 0.710 | 0.705 | +0.005 | |
148 | 0.734 | 0.731 | +0.003 | |
197 | 0.667 | 0.658 | +0.009 |
Dataset | Number of Cycles | Predicted Value | True Value | Difference |
---|---|---|---|---|
B0007 | 72 | 0.829 | 0.831 | −0.002 |
104 | 0.789 | 0.787 | +0.002 | |
136 | 0.741 | 0.739 | +0.002 | |
168 | 0.720 | 0.716 | +0.004 | |
B0033 | 68 | 0.802 | 0.812 | −0.010 |
111 | 0.700 | 0.691 | +0.009 | |
154 | 0.680 | 0.680 | 0 | |
197 | 0.667 | 0.658 | +0.009 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Liu, C.; Wang, S.; Ma, Z.; Guo, S.; Qin, Y. A Multi-Encoder BHTP Autoencoder for Robust Lithium Battery SOH Prediction Under Small-Sample Scenarios. Batteries 2025, 11, 180. https://doi.org/10.3390/batteries11050180
Liu C, Wang S, Ma Z, Guo S, Qin Y. A Multi-Encoder BHTP Autoencoder for Robust Lithium Battery SOH Prediction Under Small-Sample Scenarios. Batteries. 2025; 11(5):180. https://doi.org/10.3390/batteries11050180
Chicago/Turabian StyleLiu, Chang, Shunli Wang, Zhiqiang Ma, Siyuan Guo, and Yixiong Qin. 2025. "A Multi-Encoder BHTP Autoencoder for Robust Lithium Battery SOH Prediction Under Small-Sample Scenarios" Batteries 11, no. 5: 180. https://doi.org/10.3390/batteries11050180
APA StyleLiu, C., Wang, S., Ma, Z., Guo, S., & Qin, Y. (2025). A Multi-Encoder BHTP Autoencoder for Robust Lithium Battery SOH Prediction Under Small-Sample Scenarios. Batteries, 11(5), 180. https://doi.org/10.3390/batteries11050180