Chaotic Behavior of Battery State of Health

Batteries are used for energy storage in various applications. Renewable energy systems, such as solar cells and wind turbines, produce energy depending on the weather. Because of that, batteries are needed in these fields to ensure energy availability. Electric vehicles also rely on batteries for energy storages. Therefore, the failure of battery management can result in large losses in various aspects. To prevent this, a number of studies on battery State of Health (SoH) have been carried out. SoH Prediction and estimation of have been applied using several methods such as SVM, deep learning, random forest and others. In many cases, these methods are applied by assuming that the change in SoH is a linear or nonlinear phenomenon. However, in this paper, it is found that SoH exhibits chaotic behaviors. Several well known methods such as Autocorrelation Function, Average False Neighbors, and Lyapunov Exponent are considered to identify the chaotic nature of SoH. The obtained results can provide a new understanding of SoH behaviors on batteries, especially for the development of battery management system, control and safety systems.