An improved state of charge estimation of lithium-ion battery based on a dual input model

The state of charge (SOC) is critical for the safety and reliable utilization of the battery. However, SOC estimation accuracy of lithium iron phosphate (LiFePO4) battery is a challenging task because of its flat voltage plateau and its performance is affected greatly by temperature. An improved open circuit voltage (OCV) combined with coulomb counting (CC) method based on the dual input model of ambient temperature and OCV is proposed. This approach is employed to obtain the initial value of SOC based on the dual input model over a wide range of temperature, and to establish SOC estimation model of modified CC with Simulink. The temperature entropy coefficient and differential analysis has been analyzed under different SOC conditions. The SOC maximum errors predicted by the proposed OCV-CC method maintain within 3.10%, while the maximum errors of the traditional OCV-CC increase from 4.78% to 23.64% under the temperature of 35 degrees C similar to-15 degrees C. The results show the estimation accuracy of SOC using the proposed OCV-CC method is improved obviously, especially under low temperature.