Modeling of discharge voltage for lithium-ion batteries through orthogonal experiments at subzero environment

Lithium-ion batteries have been widely utilized in increasing number of industrial and household domains. The accuracy of the terminal voltage estimation in the discharge processes of lithium-ion batteries is crucial to ensure the availability and safety of battery-powered facilities. In prior studies, the priority of influencing factors of discharging processes, as well as the correlations between operational parameters and model parameters have not been thoroughly considered. In this paper, we conduct an orthogonal experiment of lithium-ion batteries at subzero environment. Based on experiment data, we propose the cubic polynomial to estimate the discharge voltage for lithium-ion batteries at the subzero environment. In our battery experiment, the influencing factors are the charge current, end-of-charge voltage, discharge current, and end-of-discharge voltage. We determine crucial operational parameters in the orthogonal experiment. We propose the empirical equations to depict the relationships between operational parameters and curve-fitting parameters. The results show that the cubic polynomial can be utilized for estimating and predicting the discharge voltage explicitly. We also compare the cubic polynomial with several simplified regression and machine learning methods. It is demonstrated that the former can obtain comparable results, with fewer computation resources, than the laters. In addition, the cubic polynomial also gains better explainability than the laters.