Single particle model (SPM) is an electrochemical model which can be used with energy balance to calculate the state of charge, terminal voltage, and temperature of the battery. The conventional SPM neglects the lithium-ion dynamics in the electrolyte, and the energy balance usually does not consider the heat transfer delay from the battery center to the surface. These lead to model errors when SPM is used for high-capacity battery cells. In this study, we propose several strategies to improve the model accuracy for high-capacity battery cells. First, the actual discharge capacity is considered by modifying the desired Li-ion flux at the electrode surface. Second, a part of the equivalent circuit model is integrated with SPM to consider the resistance in the electrolyte computationally efficiently. Third, the delay in the heat transfer from the center to the surface is represented using a second-order system dynamic. In electric vehicles, battery cells are stacked as a pack; we account for the additional heat generated by stacking pressure-induced swelling repression. Finally, parameter estimation is conducted to determine the best parameter values of the model using the experimental data. The reference values are set using a dimensionless scale-up approach. When the parameters are estimated and the model is tested with the validation data, the error percentages in the calculated terminal voltage and temperature are smaller than 2.0410% and 3.5032%, respectively, even with the cell variances.

单粒子模型（SPM）是一种电化学模型，可以与能量平衡一起使用来计算电池的荷电状态、端电压和温度。传统的SPM忽略了电解液中的锂离子动力学，能量平衡通常不考虑从电池中心到表面的热传递延迟。当 SPM 用于高容量电池时，这些会导致模型错误。在这项研究中，我们提出了几种策略来提高高容量电池的模型精度。首先，通过修改电极表面所需的锂离子通量来考虑实际放电容量。其次，将等效电路模型的一部分与 SPM 集成，以有效地计算电解质中的电阻。第三，从中心到表面的热传递延迟使用二阶系统动力学表示。在电动汽车中，电池单元堆叠成一个包；我们考虑了堆叠压力引起的膨胀抑制所产生的额外热量。最后，进行参数估计，利用实验数据确定模型的最佳参数值。使用无量纲放大方法设置参考值。当估计参数并使用验证数据测试模型时，计算的端电压和温度的误差百分比分别小于 2.0410% 和 3.5032%，即使存在电池差异。我们考虑了堆叠压力引起的膨胀抑制所产生的额外热量。最后，进行参数估计，利用实验数据确定模型的最佳参数值。使用无量纲放大方法设置参考值。当估计参数并使用验证数据测试模型时，计算的端电压和温度的误差百分比分别小于 2.0410% 和 3.5032%，即使存在电池差异。我们考虑了堆叠压力引起的膨胀抑制所产生的额外热量。最后，进行参数估计，利用实验数据确定模型的最佳参数值。使用无量纲放大方法设置参考值。当估计参数并使用验证数据测试模型时，计算的端电压和温度的误差百分比分别小于 2.0410% 和 3.5032%，即使存在电池差异。