锂离子电池外部短路电压行为预测模型评价

The external short circuit of lithium-ion battery produces large current, which further leads to frequent safety accidents.An accurate model is the basis of risk early warning. Focusing on the external short-circuit faults. The voltage prediction accuracy evaluation and complexity comparison of Thevenin equivalent circuit model and single particle model electrochemical model are carried out from different aspects: different initial state of charge(SOC) and different temperatures. The results show that the accuracy of Thevenin model falls with the decrease of SOC and rises with the increase of temperature. The accuracy of single particle model is less affected by initial SOC, which first rises and then falls with the increase of temperature. Aiming at the problems of the accuracy decline of Thevenin model in short circuit, a model optimization method using inductance element is proposed. The root mean square error (RMSE) of Thevenin model is less than 60 mV, and the accuracy is improved by 76%. In order to solve the problem of poor accuracy of single particle model in the external short circuit high current, a model optimization method based on electric double-layer discharge and lithium ion diffusion limitation is proposed. The RMSE of the model is less than 40 mV and the accuracy is improved by 64%. The optimized model analysis results show that the Thevenin model has high real-time performance, and the single particle model makes up for the failure of Thevenin model caused by battery leakage, but the calculation is complex. The former can be used in the early diagnosis and early warning of external short circuit, and the latter can be used in the thermal management or thermal safety research after external short circuit.