A Secure Fast Charging Control Based on A Machine Learning-Aided Electrothermal Model for Lithiumion Batteries

Abstract

Temperature is a critical safety index of Lithium-ion (Li-ion) batteries. In this paper, a dual-loop temperature-current control based on a highly coupled electrothermal model is proposed to achieve fast charging for Li-ion batteries while maintaining the battery temperature within the safety limits. The proposed electrothermal model comprises an electrical part and a thermal part. The electrical parameters in the electrical part are affected by both the state-of-charge (SoC) of the battery and the battery temperature. The generated heat source in the thermal part is affected by the open-circuit voltage, charging current, and resistive components in the electrical part. Besides, an extreme gradient boosting (XGBoost) algorithm is employed to adjust the electrical parameters against unknown disturbance over wide operating conditions. Experimental results verify the accuracy of the proposed electrothermal model in temperature prediction and validate faster charging performance of the proposed control over the conventional constant current (CC) control for a flat Lithium Polymer battery Pokefi SY300-1 within the temperature limits.