Lityum İyon Pillerin Piller Arası Sağlık Durumu Tahmini: Tek Pilin Devre Dışı Bırakılması Yöntemi ve Makine Öğrenmesi Kullanımı

Abstract

Elektrikli araçlarda ve yenilenebilir enerji depolama sistemlerinde kullanılan lityum iyon pillerin güvenliğini, güvenilirliğini ve operasyonel verimliliğini artırmak için doğru sağlık durumu (SOH) tahmini şarttır. SOH tahmini için çok sayıda makine öğrenme yaklaşımı önerilmiş olsa da, birçok çalışma performanslarını rastgele bölünmüş eğitim ve test veri kümeleri kullanarak değerlendirmektedir; bu da geliştirilen modellerin genelleme yeteneğini abartabilir. Bu çalışma, NASA pil bozulma veri kümesini kullanarak Tek Pil Dışında Bırakma (LOBO) doğrulama yöntemine dayalı, piller arası bir SOH tahmin çerçevesi sunmaktadır. Pil bozulma davranışını karakterize etmek için voltaj, akım, sıcaklık, deşarj süresi ve enerji ölçümlerinden yirmi bir istatistiksel ve fiziksel ilhamlı özellikten oluşan kapsamlı bir set çıkarılmıştır. Dört regresyon modeli, yani Rastgele Orman (RF), En Küçük Kareler Artırma (LSBoost), Uzun Kısa Süreli Bellek (LSTM) ve Destek Vektör Regresyonu (SVR), LOBO doğrulama stratejisi altında eğitilmiş ve değerlendirilmiştir; burada her pil bir kez görülmemiş bir test pili olarak kullanılırken, geri kalan piller eğitim için kullanılmıştır. Model performansı, Kök Ortalama Kare Hatası (RMSE), Ortalama Mutlak Hata (MAE) ve belirleme katsayısı (R²) kullanılarak değerlendirilmiştir. Deneysel sonuçlar, RF modelinin ortalama RMSE'si 0,0342, ortalama MAE'si 0,0305 ve ortalama R²'si 0,8654 ile en iyi genel performansı elde ettiğini göstermiştir. LSBoost ve LSTM rekabetçi tahmin doğruluğu üretirken, SVR negatif ortalama R² değeri ile zayıf piller arası genelleme sergilemiştir. Ayrıca, özellik önem analizi, VoltajRMS, Ortalama Voltaj, Enerji, Ortalama Güç ve Deşarj Süresinin SOH tahminini etkileyen en etkili değişkenler olduğunu ortaya koymuştur. Önerilen çerçeve, gerçekçi çalışma koşulları altında makine öğrenimi modellerini değerlendirmek için pratik ve güvenilir bir metodoloji sağlar ve sağlam pil sağlığı tahmin sistemleri geliştirmek için piller arası doğrulamanın önemini vurgular.

Keywords

• Lityum iyon pil
• SOH tahmini
• Tek pili dışarıda bırakma doğrulaması
• Piller arası tahmin
• Makine öğrenimi
• Özellik önemi.

Cross-Battery State-of-Health Estimation of Lithium-Ion Batteries Using Leave-One-Battery-Out Validation and Machine Learning

Abstract

Accurate state-of-health (SOH) estimation is essential for improving the safety, reliability, and operational efficiency of lithium-ion batteries used in electric vehicles and renewable energy storage systems. Although numerous machine learning approaches have been proposed for SOH prediction, many studies evaluate their performance using randomly divided training and testing datasets, which may overestimate the generalization capability of the developed models. This study presents a cross-battery SOH estimation framework based on Leave-One-Battery-Out (LOBO) validation using the NASA battery degradation dataset. A comprehensive set of twenty-one statistical and physics-inspired features was extracted from voltage, current, temperature, discharge time, and energy measurements to characterize battery degradation behavior. Four regression models, namely Random Forest (RF), Least Squares Boosting (LSBoost), Long Short-Term Memory (LSTM), and Support Vector Regression (SVR), were trained and evaluated under the LOBO validation strategy, where each battery was used once as an unseen test battery while the remaining batteries were employed for training. Model performance was assessed using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and the coefficient of determination (R²). The experimental results demonstrated that the RF model achieved the best overall performance with an average RMSE of 0.0342, an average MAE of 0.0305, and an average R² of 0.8654. LSBoost and LSTM produced competitive prediction accuracies, whereas SVR exhibited poor cross-battery generalization with a negative average R² value. Furthermore, feature importance analysis revealed that VoltageRMS, MeanVoltage, Energy, MeanPower, and DischargeTime were the most influential variables affecting SOH estimation. The proposed framework provides a practical and reliable methodology for evaluating machine learning models under realistic operating conditions and highlights the importance of cross-battery validation for developing robust battery health prediction systems.

Keywords

• Lithium-ion battery
• SOH estimation
• Leave-One-Battery-Out validation
• Cross-battery prediction
• Machine learning
• Feature importance.

References

1. [1] R. Xiong, L. Li, and J. Tian, “Towards a smarter battery management system: A critical review on battery state of health monitoring methods,” Journal of Power Sources, vol. 405, pp. 18–29, 2018.
2. [2] A. Barré, B. Deguilhem, S. Grolleau, M. Gérard, F. Suard, and D. Riu, “A review on lithium-ion battery ageing mechanisms and estimations for automotive applications,” Journal of Power Sources, vol. 241, pp. 680–689, 2013.
3. [3] W. Waag, C. Fleischer, and D. U. Sauer, “Critical review of the methods for monitoring of lithium-ion batteries in electric and hybrid vehicles,” Journal of Power Sources, vol. 258, pp. 321–339, 2014.
4. [4] X. Hu, H. Yuan, C. Zou, Z. Li, and L. Zhang, “Co-estimation of state of charge and state of health for lithium-ion batteries based on fractional-order calculus,” IEEE Transactions on Vehicular Technology, vol. 67, no. 11, pp. 10319–10329, 2018.
5. [5] M. Berecibar, I. Gandiaga, I. Villarreal, N. Omar, J. Van Mierlo, and P. Van den Bossche, “Critical review of state of health estimation methods of Li-ion batteries for real applications,” Renewable and Sustainable Energy Reviews, vol. 56, pp. 572–587, 2016.
6. [6] K. Tang, B. Luo, D. Chen, C. Wang, L. Chen, F. Li, et al., “The state of health estimation of lithium-ion batteries: A review of health indicators, estimation methods, development trends and challenges,” World Electric Vehicle Journal, vol. 16, no. 8, p. 429, 2025.
7. [7] K. Movassagh, A. Raihan, B. Balasingam, and K. Pattipati, “A critical look at coulomb counting approach for state of charge estimation in batteries,” Energies, vol. 14, no. 14, p. 4074, 2021.
8. [8] C. Vidal, P. Malysz, P. Kollmeyer, and A. Emadi, “Machine learning applied to electrified vehicle battery state of charge and state of health estimation: State-of-the-art,” IEEE Access, vol. 8, pp. 52796–52814, 2020.

9. [9] G. L. Plett, “Extended Kalman filtering for battery management systems of LiPB-based HEV battery packs: Part 3. State and parameter estimation,” Journal of Power Sources, vol. 134, no. 2, pp. 277–292, 2004.
10. [10] S. Patil, S. M. Havaldar, S. Mathad, and K. R. Patil, “Lithium-ion battery state of health estimation using support vector regression (SVR),” in Proc. 2023 Int. Conf. Ambient Intell., Knowl. Informat. Ind. Electron. (AIKIIE), 2023, pp. 1–6.
11. [11] Y. Li, X. Huang, J. Meng, K. Shi, R. Teodorescu, and D. I. Stroe, “State of health estimation for lithium-ion battery based on sample transfer learning under current pulse test,” Batteries, vol. 10, no. 5, p. 156, 2024.
12. [12] X. Sui, S. He, S. B. Vilsen, J. Meng, R. Teodorescu, and D. I. Stroe, “A review of non-probabilistic machine learning-based state of health estimation techniques for Lithium-ion battery,” Applied Energy, vol. 300, p. 117346, 2021.
13. [13] Z. Zhou, A. Aitio, and D. Howey, “Learning Li-ion battery health and degradation modes from data with aging-aware circuit models,” Applied Energy, vol. 397, p. 126375, 2025.
14. [14] R. R. Richardson, C. R. Birkl, M. A. Osborne, and D. A. Howey, “Gaussian process regression for in situ capacity estimation of lithium-ion batteries,” IEEE Transactions on Industrial Informatics, vol. 15, no. 1, pp. 127–138, 2018.
15. [15] E. Chemali, P. J. Kollmeyer, M. Preindl, R. Ahmed, and A. Emadi, “Long short-term memory networks for accurate state-of-charge estimation of Li-ion batteries,” IEEE Transactions on Industrial Electronics, vol. 65, no. 8, pp. 6730–6739, 2017.
16. [16] Q. Yao, X. Song, and W. Xie, “State of health estimation of lithium-ion battery based on CNN–WNN–WLSTM,” Complex & Intelligent Systems, vol. 10, no. 2, pp. 2919–2936, 2024.
17. [17] Y. Wang, S. Guo, Y. Cui, L. Deng, L. Zhao, J. Li, and Z. Wang, “A comprehensive review of machine learning-based state of health estimation for lithium-ion batteries: data, features, algorithms, and future challenges,” Renewable and Sustainable Energy Reviews, vol. 224, p. 116125, 2025.
18. [18] S. Wang, B. Hu, Q. Zhang, C. Fernandez, and J. M. Guerrero, “Hybrid models for predicting the state of health of lithium-ion batteries: A comprehensive review,” Journal of Energy Storage, vol. 157, p. 121600, 2026.
19. [19] J. Guirguis and R. Ahmed, “Transformer-based deep learning models for state of charge and state of health estimation of li-ion batteries: A survey study,” Energies, vol. 17, no. 14, p. 3502, 2024.
20. [20] Z. Li, X. Zhang, and W. Gao, “State of health estimation of lithium-ion battery during fast charging process based on BiLSTM-Transformer,” Energy, vol. 311, p. 133418, 2024.
21. [21] Y. Zhang, Z. Li, L. Kong, H. Xu, H. Shen, and M. Chen, “Multi-step state of health prediction of lithium-ion batteries based on multi-feature extraction and improved Transformer,” Journal of Energy Storage, vol. 105, p. 114538, 2025.
22. [22] W. He, N. Williard, M. Osterman, and M. Pecht, “Prognostics of lithium-ion batteries based on Dempster–Shafer theory and the Bayesian Monte Carlo method,” Journal of Power Sources, vol. 196, no. 23, pp. 10314–10321, 2011.
23. [23] K. Maher and N. Yerken, “Comprehensive machine learning for lithium-ion battery state-of-health estimation using group-wise cross-validation,” in Proc. 2025 14th Int. Conf. Renew. Energy Res. Appl. (ICRERA), 2025, pp. 1465–1468.
24. [24] G. R. Sylvestrin, J. N. Maciel, M. L. M. Amorim, J. P. Carmo, J. A. Afonso, S. F. Lopes, and O. H. Ando Junior, “State of the art in electric batteries’ state-of-health (soh) estimation with machine learning: A review,” Energies, vol. 18, no. 3, p. 746, 2025.
25. [25] Y. Li, K. Liu, A. M. Foley, A. Zülke, M. Berecibar, E. Nanini-Maury, et al., “Data-driven health estimation and lifetime prediction of lithium-ion batteries: A review,” Renewable and Sustainable Energy Reviews, vol. 113, p. 109254, 2019.
26. [26] D. Roman, S. Saxena, V. Robu, M. Pecht, and D. Flynn, “Machine learning pipeline for battery state-of-health estimation,” Nature Machine Intelligence, vol. 3, no. 5, pp. 447–456, 2021.
27. [27] J. Lu, R. Xiong, J. Tian, C. Wang, and F. Sun, “Deep learning to estimate lithium-ion battery state of health without additional degradation experiments,” Nature Communications, vol. 14, no. 1, p. 2760, 2023.
28. [28] K. Q. Zhou, Y. Qin, and C. Yuen, “Graph neural network-based lithium-ion battery state of health estimation using partial discharging curve,” Journal of Energy Storage, vol. 100, p. 113502, 2024.
29. [29] R. Pandit and N. Ahlawat, “A standardized comparative framework for machine learning techniques in lithium-ion battery state of health estimation,” Future Batteries, vol. 7, p. 100099, 2025.
30. [30] J. Wu, Z. Liu, Y. Zhang, D. Lei, and Y. Zhang, “A model to estimate the state of health of lithium-ion batteries for different charging rates,” Future Batteries, vol. 6, p. 100074, 2025.
31. [31] M. F. Ng, J. Zhao, Q. Yan, G. J. Conduit, and Z. W. Seh, “Predicting the state of charge and health of batteries using data-driven machine learning,” Nature Machine Intelligence, vol. 2, no. 3, pp. 161–170, 2020.
32. [32] Y. Yang, Y. Xu, Y. Nie, J. Li, S. Liu, L. Zhao, et al., “Deep transfer learning enables battery state of charge and state of health estimation,” Energy, vol. 294, p. 130779, 2024.
33. [33] Q. Li and W. Xue, “A review of feature extraction toward health state estimation of lithium-ion batteries,” Journal of Energy Storage, vol. 112, p. 115453, 2025.
34. [34] H. Li, X. Xie, X. Zhang, A. F. Burke, and J. Zhao, “Battery state estimation for electric vehicles: Translating AI innovations into real-world solutions,” Journal of Energy Storage, vol. 115, p. 116000, 2025.
35. [35] B. Saha and K. Goebel, “Battery Data Set,” NASA Prognostics Data Repository, NASA Ames Research Center, Moffett Field, CA, 2007.