The expansion force is correlated with the battery's state of charge (SOC). To quantify the relationship between expansion force and SOC, this study establishes an electrochemical–mechanical coupling model to investigate the expansion force of lithium ions during charging and discharging processes, as well as the relationship between lithium concentration and stress. Simulation results show that at the end of charging, the expansion strain of anode is 0.04, while the cathode and separator exhibit compressive strains of 0.018 and 0.05, respectively, the von Mises stress of the separator is 0.06 MPa. Furthermore, the recurrent neural network, gated recurrent unit, and long short‐term memory (LSTM) algorithms are employed to estimate the SOC of battery, comparing the prediction accuracy with or without expansion force features. Incorporating expansion force features significantly improves SOC estimation accuracy, especially under complex US06 driving conditions, reducing the mean absolute error of the LSTM model to 0.43, a 62% reduction compared to models that do not include these features. The results of this study contribute to a better understanding of battery expansion behavior and enhance both SOC estimation accuracy and battery safety.
Zhang et al. (Sun,) studied this question.