State of Charge Evaluation of Power Battery Pack Through Multi-Parameter Optimization

Abstract

The state of charge (SoC) is an important index of the energy output performance of power battery pack. But the SoC value is affected by various factors, namely, ambient temperature, working current, equilibrium potential, and the consistency between batteries in the pack. These factors might dampen the accuracy of the traditional SoC evaluation methods like current–voltage method and Kalman filter. The evaluation accuracy is also influenced by the data drift and rest time to equilibrium potential. Considering the multiple influencing factors of SoC, this paper analyzes the data drift and rest time to equilibrium potential, and builds an approximate model of overpotential for 32650 LiFePO4 battery, based on the time variation constant and the monotonicity of SoC trend. The proposed model was adopted to optimize the evaluation of SoC. To verify its effectiveness, the proposed method was compared with current–voltage method and Kalman filter through experiments. The results show that our method outperformed the contrastive methods in simplicity, relative error (<2.33%), compatibility, and state of health (SoH).