On the issue of identifying the parameters of the complete model of lithium-ion batteries obtained through the method of mathematical prototyping of energy processes
https://doi.org/10.26467/2079-0619-2025-28-6-37-52
Abstract
This paper examines the problem of identifying the parameters of a complete mathematical model of lithium-ion batteries (LIBs), based on the Method of Mathematical Prototyping of Energy Processes (MMPEP). The relevance of this topic is due to the increasing use of LIBs in aviation, including unmanned aerial systems, and the necessity to ensure the reliability and durability of batteries through accurate prediction of their characteristics. The MMPEP approach is outlined, which makes it possible to obtain models that rigorously comply with the laws of energy conservation and thermodynamics, while also considering the physicochemical characteristics of specific batteries. Particular focus is given to the stages of model parameter identification – from initial approximation based on experimental data to further optimization using modern numerical methods and machine learning algorithms. The study analyzes current tools for parameter identification, including XGBoost, Random Forest, and neural networks. It describes the development and training of an inverse neural network on synthetic data generated from the complete LIB model, and highlights the features of preparing and selecting strategies to improve prediction quality. A sensitivity analysis of the model to the various parameters is conducted, thereby enabling more targeted identification and improving the accuracy of battery diagnostics. The neural network architecture combining time-series processing and static features is presented, along with the results of experiments predicting key LIB parameters. It is noted that the obtained neural network can be useful in the rough parameter identification stage, whereas further developments will involve more complex architectures and integration of physically informed approaches to achieve more accurate mathematical models that can serve as the basis for creating digital twins of lithium- ion batteries.
About the Author
S. I. Gavrilenkov
Moscow State Technical University of Civil Aviation
Russian Federation
Stanislav I. Gavrilenkov, Postgraduate Student, Electrical Engineering and Aviation Electrical Equipment ChairMoscow
Russian Federation
Stanislav I. Gavrilenkov, Postgraduate Student, Electrical Engineering and Aviation Electrical Equipment ChairMoscow
References
1. Ivanov, V.V., Marakhovskiy, I.V., Kravchenko, S.V. (2013). Formulation of requirements for aviation lithium-ion batteries. In: X nauchnyye chteniya, posvyashchennyye pamyati N.Ye. Zhukovskogo: materialy Vserossiyskoy nauchno-tekhnicheskoy konferentsii. Moscow: Izdatelkiy dom Akademii imeni N.E. Zhukovsky, pp. 303–306. (in Russian)
2. Kedrinskiy, I.A., Yakovlev, V.G. (2002). Lithium-ion batteries. Krasnoyarsk: Platina, 268 p. (in Russian)
3. Starostin, I.E., Stepankin, A.G. (2019). Software implementation of modern nonequilibrium thermodynamics methods and simulation system of physicochemical processes SimulationNonEqProcSS v.0.1.0: Monograph. Beau Bassin, Mauritius: Lambert Academic Publishing, 127 p. (in Russian)
4. Starostin, I.E., Khalyutin, S.P. (2024). An analytical model of lithium-ion battery voltage dynamics. Elektrichestvo, no. 10, pp. 13–22. DOI: 10.24160/0013-5380-2024-10-13-22 (in Russian)
5. Khalyutin, S.P., Starostin, I.E., Agafonkina, I.V. (2023). Generalized method of mathematical prototyping of energy processes for digital twins development. Energies, vol. 16, issue 4, ID: 1933. DOI: 10.3390/en16041933 (accessed: 10.03.2025).
6. Khalyutin, S.P., Starostin, I.E., Davidov, A.O., Kharkov, V.P., Zhmurov, B.V. (2022). Digital twins in the theory and practices of aircraft electrical power systems. Elektrichestvo, vol. 10, pp. 4–13. DOI: 10.24160/00135380-2022-10-4-13 (in Russian)
7. Starostin, I.E. (2024). Building, based on interpolation, models of various physical and chemical systems by method of mathematical prototyping of energy processes. Reliability and Quality of Complex Systems, no. 1 (45), pp. 49–58. DOI: 10.21685/2307-4205-2024-1-6 (in Russian)
8. Starostin, I.E., Druzhinin, A.A., Gavrilenkov, S.I. (2023). Application of supervised machine learning for building mathematical models of systems by the method of mathematical prototyping of energy processes. In: Trudy Mezhdunarodnogo simpoziuma “Nadezhnost i kachestvo”, vol. 1, pp. 66–72. (in Russian)
9. Starostin, I.E., Khalyutin, S.P., Davidov, A.A., Levin, A.V., Trubachev, A.I. (2019). The development of a mathematical model of lithium-ion battery discharge characteristics. In: Proceedings ICOECS 2019: 2019 international conference on electrotechnical complexes and systems, pp. 8949976. DOI: 10.1109/ICOECS46375.2019.8949976
10. Starostin, I.E., Gavrilenkov, S.I. (2024). Architecture of the mathematical core of digital twins of various physical and chemical systems based on the method of mathematical prototyping of energy processes. Reliability and Quality of Complex Systems, no. 4 (48), pp. 160–168. DOI: 10.21685/2307-4205-2024-4-17 (in Russian)
11. Starostin, I.E., Druzhinin, A.A. (2023). Analytical approximation of solutions of equations of the method of mathematical prototyping of energy processes by qualitative analysis of these equations. Reliability and Quality of Complex Systems, no. 2 (42), pp. 22–31. DOI: 10.21685/2307-4205-2023-2-3 (in Russian)
12. Starostin, I.E., Gavrilenkov, S.I. (2025). Assignment of state functions for interaction potentials, reduced heat capacities and reduced thermal effects included in the equations of the method of mathematical prototyping of energy processes. Reliability and Quality of Complex Systems, no. 1 (49), pp. 36–44. DOI: 10.21685/2307-4205-2025-1-5 (in Russian)
13. Alshawabkeh, A., Matar, M., Almutairy, F. (2024). Parameters identification for lithium-ion battery models using the LevenbergMarquardt algorithm. World Electric Vehicle Journal, vol. 15, issue 9, ID: 406. DOI: 10.3390/wevj15090406 (accessed: 26.03.2025).
14. Lian, Y., Qiao, D. (2023). A novel capacity estimation method for lithium-ion batteries based on the adam algorithm. Batteries, vol. 11, issue 3. ID: 85. DOI: 10.3390/ batteries11030085 (accessed: 26.03.2025).
15. Pi, J., da Silva, S.F., Ozkan, M.F., Gupta, A., Canova, M. (2024). Parameter identification for electrochemical models of lithiumion batteries using Bayesian optimization. IFAC-PapersOnLine, vol. 57, no. 1, pp. 180–185. DOI: 10.1016/j.ifacol.2024.12.031
16. Starostin, I.E., Druzhinin, A.A. (2023). The concept of a software and technological platform for digital twins based on energy dynamics methods. In: 2023 5th International Youth Conference on Radio Electronics, Electrical and Power Engineering (REEPE), pp. 1–6. DOI: 10.1109/REEPE57272.2023.10086710
17. Song, S., Fei, C., Xia, H. (2020). Lithium-Ion Battery SOH Estimation Based on XGBoost Algorithm with Accuracy Correction. Energies, vol. 13, issue 4, ID: 812. DOI: 10.3390/en13040812 (accessed: 26.03.2025).
18. Naaz, N., Channegowda, J. (2022). XGBoost based synthetic battery parameter generation to overcome limited battery dataset challenges. In: 2022 IEEE International Conference on Power Electronics, Smart Grid, and Renewable Energy (PESGRE). India, Trivandrum, pp. 1–4. DOI: 10.1109/PESGRE52268. 2022.9715814
19. Heinrich, F., Klapper, P., Pruckner, M. (2021). A comprehensive study on battery electric modeling approaches based on machine learning. Energy Informatics, vol. 4, ID: 17. DOI: 10.1186/s42162-021-00171-7 (accessed: 26.03.2025).
20. Singh, S., Ebongue, Y.E., Rezaei, S., Birke, K.P. (2023). Hybrid modeling of lithiumion battery: physics-informed neural network for battery state estimation. Batteries, vol. 9, issue 6, ID: 301. DOI: 10.3390/batteries9060301 (accessed: 26.03.2025).
21. Wang, J., Peng, Q., Meng, J., Liu, T., Peng, J., Teodorescu, R. (2024). A physicsinformed neural network approach to parameter estimation of lithium-ion battery electrochemical model. Journal of Power Sources, vol. 621, ID: 235271. DOI: 10.1016/j.jpowsour.2024.235271 (accessed: 26.03.2025).
22. Shen, S., Sadoughi, M., Li, M., Wang, Z., Hu, C. (2020). Deep convolutional neural networks with ensemble learning and transfer learning for capacity estimation of Lithium-Ion batteries. Applied Energy, vol. 260, ID: 114244. DOI: 10.1016/j.apenergy.2019.114244 (accessed: 26.03.2025).
Review
For citations:
Gavrilenkov S.I. On the issue of identifying the parameters of the complete model of lithium-ion batteries obtained through the method of mathematical prototyping of energy processes. Civil Aviation High Technologies. 2025;28(6):37-52. (In Russ.) https://doi.org/10.26467/2079-0619-2025-28-6-37-52
Views:
15
JATS XML
Email this article 
