ABSTRACT Lithium batteries are widely used in energy storage and new energy electric vehicles due to their superior performance,but the internal short circuit problem of lithium batteries is a safety hazard during use for energy storage and vehicle battery packs.If it cannot be detected in time,the deepening of the internal short circuit will be accompanied by an increase in heat,which will cause thermal runaway and lead to safety accidents.Diagnose whether the battery pack has an internal short circuit and quantitatively estimate the short circuit resistance of the battery cell that has the internal short circuit,it can effectively prevent the pecurrence of thermal runaway.Based on the charging curve of lithium-ion battery module,a quantitative diagnosis algorithm of ISC on the basis of the remaining charge capacity (RCC)is proposed in this study.The simulation and experimental verification of the algorithm are carried out under the conditions of different voltage acquisition accuracies,sampling periods,temperatures,and aging degrees.The results show that the proposed algorithm can accurately and quantitatively diagnose the ISC under certain conditions:(1)For the serious ISC of 10 level, high diagnosis accuracy can be obtained even under the conditions of 10 mV acquisition accuracy,10 s sampling period and variable temperature.For the early ISC of 100Q level,the ISC resistance is smaller than the actual value,and the diagnosis time is longer.To improve the accuracy and timeliness of early ISC diagnosis,the voltageacquisition accuracy and sampling frequency should be higherhanand H,respectively:(2) Battery aging will reduce the accuracy of ISC diagnosis,but it has little effect on 10 level ISC,and the diagnostic error of ISC resistance is less than 6%even at extreme low temperature(-20C).The conclusions are of great significance to improve the accuracy of quantitative diagnosis of ISC for lithium-ion batteries. KEY WORDS Lithium-ion batteries;Internal short cifcuit;Remaining charge capacity;Fault diagnosis; Battery safety 能源与环境问题的日益突出促进了锂电池的蓬勃发展，而锂电池直接关系到各个能 源领域的安全性及经济性。由于电池现有制造工艺的缺陷和电池使用过程中的滥用行 为（电滥用、热滥用及机械滥用） 龟池系统中的个别单体电池可能出现内短路故障5 。内短路故障一经形成， 会不断消耗电池电量并产生热量。如果内短路进一步发展可能 引发热失控等严重的安全事做， 因此， 内短路的早期诊断与预警对提高电池系统的安全 性至关重要9-13。 然而，内短路具潜伏期长隐蔽性强等特征，这给内短路的诊断带来了困难。 现有内短路在线诊断方法注要分为两类：一类是内短路的定性在线诊断方法。冯旭宁等 人根据等效电路模型辨识出的参数差异进行内短路检测。这种方法将内短路检测问题 转化成模型参数估计问题，通过建立等效电路模型，利用测得的单体电压、温度和电流 数据在线估计电池时荷电状态(State of charge,SOC)与欧姆内阻等特征参数，进而将 表现最差单体的特征参数与“平均单体”进行比较来检测内短路。另一类是内短路的定 量诊断方法。典型方法是一种基于剩余充电电量(Remaining charge capacity,RCC)变化 的内短路检测方法，基本原理是通过比较各电池连续两次剩余充电电量变化来检测内 短路。该方法无需建立模型就能够准确检测出内短路并进行量化，大大减少了BMS的计 算负载与储存空间。定量诊断方法通过计算内短路阻值来评估电池内短路程度，从而避 免电池热失控的发生。 在现有基于剩余充电电量变化的内短路检测方法中，内短路阻值的定量计算精度取 决于RCC的估计精度，而RCC的估计精度会受到电池老化程度、温度、采样精度与频率 等因素的影响。因此，研究多种因素影响下的内短路定量诊断方法对提高内短路诊断的 时效性与准确性具有重要意义。本文利用充电工况下的电量关系提出了一种内短路定量 在线诊断算法，并利用仿真与实验的方法验证该算法在多种因素影响下的有效性。 首先介绍RCC的估计原理及内短路定量诊断方法：其次，建立电池模组的内短路仿真 模型，并对不同的电压采样频率与精度下的内短路进行诊断：最后，利用电池模组实 验对不同老化程度及变温度下的内短路诊断结果进行研究与分析，研究结论对提高内短ABSTRACT Lithium batteries are widely used in energy storage and new energy electric vehicles due to their superior performance, but the internal short circuit problem of lithium batteries is a safety hazard during use for energy storage and vehicle battery packs. If it cannot be detected in time, the deepening of the internal short circuit will be accompanied by an increase in heat, which will cause thermal runaway and lead to safety accidents. Diagnose whether the battery pack has an internal short circuit and quantitatively estimate the short circuit resistance of the battery cell that has the internal short circuit, it can effectively prevent the occurrence of thermal runaway. Based on the charging curve of lithium-ion battery module, a quantitative diagnosis algorithm of ISC on the basis of the remaining charge capacity (RCC) is proposed in this study. The simulation and experimental verification of the algorithm are carried out under the conditions of different voltage acquisition accuracies, sampling periods, temperatures, and aging degrees. The results show that the proposed algorithm can accurately and quantitatively diagnose the ISC under certain conditions: (1) For the serious ISC of 10 Ω level, high diagnosis accuracy can be obtained even under the conditions of 10 mV acquisition accuracy, 10 s sampling period and variable temperature. For the early ISC of 100 Ω level, the ISC resistance is smaller than the actual value, and the diagnosis time is longer. To improve the accuracy and timeliness of early ISC diagnosis, the voltage acquisition accuracy and sampling frequency should be higher than 1mV and 1Hz, respectively; (2) Battery aging will reduce the accuracy of ISC diagnosis, but it has little effect on 10 Ω level ISC, and the diagnostic error of ISC resistance is less than 6% even at extreme low temperature (-20 ℃). The conclusions are of great significance to improve the accuracy of quantitative diagnosis of ISC for lithium-ion batteries. KEY WORDS Lithium-ion batteries; Internal short circuit; Remaining charge capacity; Fault diagnosis; Battery safety 能源与环境问题的日益突出促进了锂电池的蓬勃发展，而锂电池直接关系到各个能 源领域的安全性及经济性[1-4]。由于电池现有制造工艺的缺陷和电池使用过程中的滥用行 为（电滥用、热滥用及机械滥用），电池系统中的个别单体电池可能出现内短路故障[5- 8]。内短路故障一经形成，会不断消耗电池电量并产生热量。如果内短路进一步发展可能 引发热失控等严重的安全事故，因此，内短路的早期诊断与预警对提高电池系统的安全 性至关重要[9-13]。 然而，内短路具有潜伏期长、隐蔽性强等特征[14]，这给内短路的诊断带来了困难。 现有内短路在线诊断方法主要分为两类：一类是内短路的定性在线诊断方法。冯旭宁等 人[15]根据等效电路模型辨识出的参数差异进行内短路检测。这种方法将内短路检测问题 转化成模型参数估计问题，通过建立等效电路模型，利用测得的单体电压、温度和电流 数据在线估计电池的荷电状态（State of charge，SOC）与欧姆内阻等特征参数，进而将 表现最差单体的特征参数与“平均单体”进行比较来检测内短路。另一类是内短路的定 量诊断方法。典型方法是一种基于剩余充电电量（Remaining charge capacity，RCC）变化 的内短路检测方法[16]，基本原理是通过比较各电池连续两次剩余充电电量变化来检测内 短路。该方法无需建立模型就能够准确检测出内短路并进行量化，大大减少了 BMS 的计 算负载与储存空间。定量诊断方法通过计算内短路阻值来评估电池内短路程度，从而避 免电池热失控的发生。 在现有基于剩余充电电量变化的内短路检测方法中，内短路阻值的定量计算精度取 决于 RCC 的估计精度，而 RCC 的估计精度会受到电池老化程度、温度、采样精度与频率 等因素的影响。因此，研究多种因素影响下的内短路定量诊断方法对提高内短路诊断的 时效性与准确性具有重要意义。本文利用充电工况下的电量关系提出了一种内短路定量 在线诊断算法，并利用仿真与实验的方法验证该算法在多种因素影响下的有效性。 首先介绍 RCC 的估计原理及内短路定量诊断方法；其次，建立电池模组的内短路仿真 模型，并对不同的电压采样频率与精度下的内短路进行诊断；最后，利用电池模组实 验对不同老化程度及变温度下的内短路诊断结果进行研究与分析，研究结论对提高内短 录用稿件，非最终出版稿