Using machine learning for state of charge estimation in high poker batteries in the defense ındustry

Lithium-ion batteries are widely preferred in energy storage systems due to their high energy density, lightweight structure, and long lifespan. These batteries are utilized in various fields such as the transportation sector, consumer electronics, and renewable energy applications. However, the accurate estimation of the State of Charge (SoC) is of great importance to ensure the efficient and safe utilization of these batteries. In this study, a dataset obtained from eight lithium-ion batteries subjected to 500 charge-discharge cycles under different operating conditions was used. The dataset includes parameters such as battery temperature, applied discharge current, and the number of charge-discharge cycles. During the data preprocessing phase, missing values were handled, outliers were identified, and feature engineering tasks were carried out. Various machine learning algorithms were applied to the dataset. Among the implemented machine learning methods, the Random Forest Regressor model was found to be the most effective in achieving the highest prediction performance.