Estimated State of Charge of Li-Ion Batteries Using GRU Deep Learning
Keywords:
Deep Learning, Gated Recurrent Unit, Li-ion Battery, Renewable Energy, State of ChargeAbstract
State of Charge (SoC) estimation is one of the most important functions of a Battery Management System (BMS) because it directly affects the performance, lifespan, and operational safety of lithium-ion batteries, particularly in renewable energy applications with dynamic operating conditions. Conventional estimation methods are susceptible to cumulative integration errors and performance degradation under varying ambient temperatures. This study proposes a Gated Recurrent Unit (GRU)-based deep learning model for SoC estimation using five input features: voltage (V), current (I), temperature (T), moving average voltage (Vavg), and moving average current (Iavg). The additional moving-average features were introduced to enhance the temporal representation of battery behavior and improve estimation accuracy. Experimental data from the LG 18650 HG2 battery dataset were used under four operating temperatures (0°C, 10°C, 25°C, and 40°C). The proposed GRU model was evaluated using different training epochs, where optimal performance was achieved at 2,400 epochs, resulting in a validation RMSE of 1.42%. Furthermore, the proposed five-input configuration reduced the validation RMSE from 4.34% to 1.43% at 1,200 epochs, corresponding to an improvement of 67.05% compared with the conventional three-input configuration. These results demonstrate that incorporating moving-average features significantly improves the estimation accuracy and robustness of the GRU model under varying temperature conditions, providing an effective data-driven approach for lithium-ion battery SoC estimation.
Abstrak: Estimasi State of Charge (SoC) merupakan salah satu fungsi paling penting dalam Battery Management System (BMS) karena berpengaruh langsung terhadap performa, masa pakai, dan keselamatan operasional baterai litium-ion, khususnya pada aplikasi energi terbarukan dengan kondisi operasi yang dinamis. Metode estimasi konvensional rentan terhadap akumulasi galat integrasi dan penurunan performa di bawah temperatur lingkungan yang bervariasi. Penelitian ini mengusulkan model deep learning berbasis Gated Recurrent Unit (GRU) untuk estimasi SoC menggunakan lima fitur masukan: tegangan (V), arus (I), temperatur (T), tegangan rata-rata (Vavg), dan arus rata-rata (Iavg). Fitur rata-rata berjalan tambahan tersebut diperkenalkan untuk meningkatkan representasi temporal dari perilaku baterai dan memperbaiki akurasi estimasi. Data eksperimental dari dataset baterai LG 18650 HG2 digunakan di bawah empat temperatur operasi (0°C, 10°C, 25°C, dan 40°C). Model GRU yang diusulkan dievaluasi menggunakan jumlah epoch pelatihan yang berbeda, di mana performa optimal dicapai pada 2.400 epoch, menghasilkan RMSE validasi sebesar 1,42%. Lebih lanjut, konfigurasi lima masukan yang diusulkan berhasil menurunkan RMSE validasi dari 4,34% menjadi 1,43% pada 1.200 epoch, yang selaras dengan peningkatan sebesar 67,05% dibandingkan dengan konfigurasi tiga masukan konvensional. Hasil ini menunjukkan bahwa integrasi fitur rata-rata berjalan secara signifikan meningkatkan akurasi estimasi dan ketahanan model GRU di bawah kondisi temperatur yang bervariasi, serta menyediakan pendekatan berbasis data (data-driven) yang efektif untuk estimasi SoC baterai litium-ion.
References
[1] A. E. Setyono and B. F. T. Kiono, “Dari Energi Fosil Menuju Energi Terbarukan: Potret Kondisi Minyak dan Gas Bumi Indonesia Tahun 2020 – 2050,” J. Energi Baru dan Terbarukan, vol. 2, no. 3, pp. 154–162, Oct. 2021, doi: https://doi.org/10.14710/jebt.2021.11157.
[2] M. U. Ali, A. Zafar, S. H. Nengroo, S. Hussain, M. Junaid Alvi, and H.-J. Kim, “Towards a Smarter Battery Management System for Electric Vehicle Applications: A Critical Review of Lithium-Ion Battery State of Charge Estimation,” Energies, vol. 12, no. 3. p. 446, 2019. doi: https://doi.org/10.3390/en12030446.
[3] M. A. Hannan, M. S. H. Lipu, A. Hussain, and A. Mohamed, “A review of lithium-ion battery state of charge estimation and management system in electric vehicle applications: Challenges and recommendations,” Renew. Sustain. Energy Rev., vol. 78, pp. 834–854, 2017, doi: https://doi.org/10.1016/j.rser.2017.05.001.
[4] Y. Zhang, R. Xiong, H. He, and W. Shen, “Lithium-Ion Battery Pack State of Charge and State of Energy Estimation Algorithms Using a Hardware-in-the-Loop Validation,” IEEE Trans. Power Electron., vol. 32, no. 6, pp. 4421–4431, 2017, doi: https://doi.org/10.1109/TPEL.2016.2603229.
[5] A. A. Husien R, N. A. Windarko, and B. Sumantri, “Estimasi State of Charge (SOC) Pada Baterai Lithium Ion Menggunakan Long Short-Term Memory (LSTM) Neural Network,” Briliant J. Ris. dan Konseptual, vol. 9, no. 4, pp. 986–997, 2024, doi: https://doi.org/10.28926/briliant.v9i4.1955.
[6] H. He, R. Xiong, X. Zhang, F. Sun, and J. Fan, “State-of-Charge Estimation of the Lithium-Ion Battery Using an Adaptive Extended Kalman Filter Based on an Improved Thevenin Model,” IEEE Trans. Veh. Technol., vol. 60, no. 4, pp. 1461–1469, 2011, doi: https://doi.org/10.1109/TVT.2011.2132812.
[7] Z. Deng, X. Hu, X. Lin, Y. Che, L. Xu, and W. Guo, “Data-driven state of charge estimation for lithium-ion battery packs based on Gaussian process regression,” Energy, vol. 205, p. 118000, 2020, doi: https://doi.org/10.1016/j.energy.2020.118000.
[8] L. Alzubaidi et al., “Review of deep learning: concepts, CNN architectures, challenges, applications, future directions,” J. Big Data, vol. 8, no. 1, p. 53, 2021, doi: https://doi.org/10.1186/s40537-021-00444-8.
[9] D. E. Cahyani, L. Gumilar, A. N. Afandi, A. P. Wibawa, and A. K. Junoh, “Hybrid deep learning for estimation of state-of-health in lithium-ion batteries,” Int. J. Electr. Comput. Eng., vol. 15, no. 1, pp. 995–1006, Feb. 2025, doi: http://doi.org/10.11591/ijece.v15i1.pp995-1006.
[10] J. Chen, Y. Zhang, W. Li, W. Cheng, and Q. Zhu, “State of charge estimation for lithium-ion batteries using gated recurrent unit recurrent neural network and adaptive Kalman filter,” J. Energy Storage, vol. 55, p. 105396, 2022, doi: https://doi.org/10.1016/j.est.2022.105396.
[11] M. Catelani, L. Ciani, R. Fantacci, G. Patrizi, and B. Picano, “Remaining Useful Life Estimation for Prognostics of Lithium-Ion Batteries Based on Recurrent Neural Network,” IEEE Trans. Instrum. Meas., vol. 70, pp. 1–11, 2021, doi: https://doi.org/10.1109/TIM.2021.3111009.
[12] H. Chaoui and C. C. Ibe-Ekeocha, “State of Charge and State of Health Estimation for Lithium Batteries Using Recurrent Neural Networks,” IEEE Trans. Veh. Technol., vol. 66, no. 10, pp. 8773–8783, 2017, doi: https://doi.org/10.1109/TVT.2017.2715333.
[13] R. Rakhmawati, Sutedjo, F. N. Oktaviani, Irianto, D. S. Yanaratri, and A. F. Adila, “Estimasi State of Charge pada Baterai Lead Acid menggunakan Elman Recurrent Neural Network,” ELKOMIKA J. Tek. Energi Elektr. Tek. Telekomun. Tek. Elektron., vol. 11, no. 4, pp. 864–878, Oct. 2023, doi: https://doi.org/10.26760/elkomika.v11i4.864.
[14] G. Van Houdt, C. Mosquera, and G. Nápoles, “A review on the long short-term memory model,” Artif. Intell. Rev., vol. 53, no. 8, pp. 5929–5955, 2020, doi: https://doi.org/10.1007/s10462-020-09838-1.
[15] C. B. Vennerød, A. Kjærran, and E. S. Bugge, “Long Short-term Memory RNN.” 2021. doi: https://doi.org/10.48550/arXiv.2105.06756.
[16] S. Jafari and Y.-C. Byun, “A CNN-GRU Approach to the Accurate Prediction of Batteries’ Remaining Useful Life from Charging Profiles,” Computers, vol. 12, no. 11. p. 219, 2023. doi: https://doi.org/10.3390/computers12110219.
[17] S. Cui and I. Joe, “A Dynamic Spatial-Temporal Attention-Based GRU Model With Healthy Features for State-of-Health Estimation of Lithium-Ion Batteries,” IEEE Access, vol. 9, pp. 27374–27388, 2021, doi: https://doi.org/10.1109/ACCESS.2021.3058018.
[18] C. Li, F. Xiao, and Y. Fan, “An Approach to State of Charge Estimation of Lithium-Ion Batteries Based on Recurrent Neural Networks with Gated Recurrent Unit,” Energies, vol. 12, no. 9. p. 1592, 2019. doi: https://doi.org/10.3390/en12091592.
[19] Y. Gao, C. Zhu, X. Zhang, and B. Guo, “Implementation and evaluation of a practical electrochemical- thermal model of lithium-ion batteries for EV battery management system,” Energy, vol. 221, p. 119688, 2021, doi: https://doi.org/10.1016/j.energy.2020.119688.
[20] Y. Wang, X. Han, D. Guo, L. Lu, Y. Chen, and M. Ouyang, “Physics-Informed Recurrent Neural Networks with Fractional-Order Constraints for the State Estimation of Lithium-Ion Batteries,” Batteries, vol. 8, no. 10. p. 148, 2022. doi: https://doi.org/10.3390/batteries8100148.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Muhammad Rezki, Novie Ayub Windarko, Bambang Sumantri
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
