This study proposed a dynamic resistance-based thermal model to predict the temperature evolution of a prismatic lithium-ion battery in fast and regular charging strategies. The effects of battery temperature, state of charge (SOC), and charging current on resistance were investigated to form a dynamic heat generation model. This model included both the heat generation of battery body and the heat generation of electrodes. Then the charging resistance-based thermal model was used to predict the temperature distribution and temperature evolution of a 50 Ah prismatic battery under different charging strategies and ambient temperatures. The charging strategies included a multi-stage constant current-constant voltage (MSCC-CV) for fast charging and a constant current-constant voltage (CC-CV) for regular charging. Experiments verified the prediction accuracy of the proposed thermal model. The results showed that the dynamic thermal model accurately predicted the temperature distribution and its evolution under different charging strategies and ambient temperatures. With regard to temperature distribution prediction, the average root mean square errors (RMSEs) of temperature among the five test points were 0.43 °C for the MSCC-CV charging strategy and 0.3 °C for the CC-CV charging strategy. For the temperature prediction under various ambient temperatures, the average RMSEs at different ambient temperatures were 0.34 °C for the MSCC-CV strategy and 0.25 °C for the CC-CV strategy. Finally, the effect of the charging and discharging resistance on the temperature prediction under the MSCC-CV charging condition was studied. Although the discharging resistance-based thermal model could describe the temperature change with charging time, it showed higher predicted temperatures than the tested values and a larger estimation error than the charging resistance-based thermal model.
Read full abstract