(Invited) Parameter Identification and Degradation Estimation of Li-Ion Batteries: Physics-Based Model and EIS Experimental Coupling

Lithium-Ion Battery (LIB) diagnosis has generally been handled via two opposite approaches: academic approach vs industrial deployed approach [1]. The academic approach [2, 3] focuses more on prediction accuracy in the expense of destruction, high man-hour, expensive experiments, and heavy computing; the industrial deployed approach [4] on the contrary uses little resources and is nondestructive but is often restricted to an extrapolation of the evolution of capacity and resistance, which becomes ineffective in predicting sudden acceleration of capacity fade and battery status. An intermediate route to reach an accurate diagnosis with a cost-effective and nondestructive method is lacking. On the other hand, a comprehensive battery evaluation approach is unavailable. Most manufacturers only provide the state of charge (SOC) as an indicator of the battery status. There are three battery status measures[5]: SOC representing the remaining charge of the battery; state of health (SOH) describing the extent of battery aging; state of function (SOF) describing how the battery performance meets the real load demands. To evaluate a battery status in real applications, all three measures should be co-estimated, which remains a scientific challenge, though is critical to safe operation of batteries. Electrochemical impedance spectroscopy (EIS)[6] is well-known as a comprehensive and non-destructive technique to investigate electrical properties of LIBs in association with electrochemical processes. EIS measurement is obtained in a wide frequency range, which provides the possibility to separate the internal electrochemical processes with different time constants. To interpret EIS, mathematical modeling of LIBs based on physical-electrochemical processes is more effective to capture the key characteristics of batteries. Physics-based models rely on physical laws to describe the electrochemistry inside of a battery and hence offer a better understanding of the underlying physics. They provide detailed insight in the different electrochemical phenomena involved in the battery. Another benefit of these physics-based models is that they can be coupled with other physics-based models in a consistent way, in order to extend their capabilities and incorporate other physical phenomena, such as thermal or degradation effects. However, Physics-based models are computationally expensive and require many parameters. In addition, the parameters need to be measured or determined by performing a range of experiments, which is a challenge. In this research, we will be working with in-house assembled Li(Ni0.5Mn0.3Co0.2)O2/Li-metal half-cells for EIS testing (Fig. 1a) and simulation. Because the complexity of Physics-based models, we will firstly establish a mathematical framework to evaluate the parameter identifiability through sensitivity and correlation analysis (Fig. 1b and 1c), which will help down-select a group of independent and sensitive parameters which could not be directly measured from experiments. Then, we will use the model to identify thermodynamic and kinetic parameters (Fig. 1d and 1e) for both fresh cells and degraded cells under different cycles. We will be able to map the physical parameter evolution profiles with the capacity degradation of the battery. We are also planning to develop a physics-informed machine learning model to accelerate the computational process and realize fast prediction of the cell degradation. References Dubarry, M., Battery Intelligence: Diagnosis and Prognosis. 2021, Office of Naval Research (APRISES), SAFT (France), Element Energy, ACCURE (Germany): University of Warwick, University of Oviedo, Naval Research Laboratory, Dalhousie University, Oxford University, London Imperial College, Free University of Brussels, University of Bayreuth. Verma, A., et al., Galvanostatic Intermittent Titration and Performance Based Analysis of LiNi0.5Co0.2Mn0.3O2Cathode. Journal of The Electrochemical Society, 2017. 164(13): p. A3380-A3392. Chen, C.-H., et al., Development of Experimental Techniques for Parameterization of Multi-scale Lithium-ion Battery Models. Journal of The Electrochemical Society, 2020. 167(8): p. 080534. Chang, W.-Y., The State of Charge Estimating Methods for Battery: A Review. ISRN Applied Mathematics, 2013. 2013: p. 953792. Shen, P., et al., The Co-estimation of State of Charge, State of Health, and State of Function for Lithium-Ion Batteries in Electric Vehicles. IEEE Transactions on Vehicular Technology, 2018. 67(1): p. 92-103. Orazem, M.E. and B. Tribollet, Electrochemical impedance spectroscopy. New Jersey, 2008: p. 383-389. Figure 1

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.