磷酸铁锂电池因其优异的电化学稳定性和长寿命等优点, 被广泛应用于储能系统, 但大容量储能电池内部存在电流不均匀性问题。当磷酸铁锂电池内部电流差异较大时, 电池端电压曲线形状会发生改变, 从而加速电池老化。为此, 建立准确的电池热-电-老化耦合模型有利于提高储能电池系统的长期运维性能和可靠性。该文通过改进磷酸铁锂电池等效电路模型, 模拟电池内部电流不均匀性导致的电池端电压曲线形状改变特性, 结合温度估计模型与老化模型建立了磷酸铁锂电池热-电-老化耦合模型。研究结果表明：该文所提耦合模型能够精准预测磷酸铁锂电池端电压、温度和老化状态, 除电流切换位置外, 其他区间电池端电压模拟精度绝对误差小于20.0mV, 温度估计绝对误差小于1.0℃, 电池容量衰减模拟均方根误差小于3.000Ah。该文研究结果可为储能电池系统的健康状态评估和安全管理提供参考。

Objective: The excellent electrochemical stability, high safety, and long service life of lithium iron phosphate (LiFePO₄) batteries have led to their widespread use in energy storage systems. However, in practical applications, large-capacity energy storage batteries often suffer from significant inhomogeneity of the internal current. This uneven current distribution can lead to localized temperature variations, alter the shape of the battery's terminal voltage curve, and accelerate degradation processes such as the formation of lithium dendrites and thermal stress. Existing models fail to comprehensively account for the coupled thermal-electrical-aging characteristics when modeling LiFePO₄ batteries, rendering them incapable of accurately reflecting the variations in the voltage curve caused by inconsistencies in the internal current distribution. Furthermore, such models lack credible and systematic validation across multiple operating conditions. Methods: Thus, in this study, a 280.000 Ah LiFePO₄ battery was selected as the research target for investigating a thermal-electrical-aging coupling model for large-capacity LiFePO₄ batteries that considers changes in the shape of the voltage curve. By modeling the battery as a parallel combination of multiple subcells with varying electrical characteristics, the internal inhomogeneity of the battery and the resulting influence on deformation of the voltage curve are effectively simulated, where the voltage of the parallel battery pack serves as the terminal voltage output by the model. A temperature estimation module is integrated to simulate the processes of heat generation and transfer, while an aging module is introduced to capture the evolution of capacity degradation. Together, these modules form a comprehensive thermal-electrical-aging coupling model. By constructing a comprehensive thermal-electrical-aging coupling model, the terminal voltage, temperature, and capacity of the battery can be obtained directly from the input current. The model is then validated under constant current discharge conditions with variations in temperature, dynamic operating conditions, and multiple charge-discharge cycle conditions. Results: Experimental validation using LiFePO₄ batteries demonstrated that the proposed model could accurately predict the voltage, temperature, and aging state with relatively low computational complexity. Specifically, during constant current discharging of LiFePO₄ batteries, the root mean square error (RMSE) of the temperature estimation remained below 1.0 ℃ across different temperatures and discharge rates, except under the highest rate condition. The increased RMSE at the highest discharge rate was attributed to a larger internal-external temperature gradient, which reduced the accuracy of the estimation. Additionally, faster heat dissipation at lower temperatures further reduced the precision of the temperature prediction. Because the thermal and electrical models were coupled, their respective errors were compounded. Across the four dynamic conditions, the absolute error in the voltage simulated by the model was below 20.0 mV for all intervals, except at the current-switching points, and the absolute error for the temperature estimation was below 1.0 ℃. The RMSEs for capacity degradation estimated through simulation using the coupled model were 0.223, 1.640, 1.320, and 2.700 Ah for cycling aging at 35.0 ℃/0.5 C, 35.0 ℃/1.0 C, 45.0 ℃/0.5 C, and 45.0 ℃/1.0 C, respectively—all below 1% of the total capacity—demonstrating the strong ability of the model to accurately simulate the battery capacity after aging. Conclusions: The proposed thermal-electrical-aging coupling model effectively addresses the limitations of traditional equivalent circuit models, which often lack the capability to account for inhomogeneity in the internal current, temperature variations, and aging effects. This model thus provides a solid theoretical foundation and practical methodology for estimating the state of the battery in real-time, diagnosing faults, and predicting the lifetime of energy storage systems.