随着电动汽车的大规模推广，以热失控为特征的锂离子动力电池系统安全事故时有发生，影响了电动汽车的市场接受度。动力电池热失控防范既影响电动汽车当前的市场竞争力和品牌塑造，也决定着电动汽车的未来发展，急需重点研究。 本课题针对车用锂离子动力电池全生命周期热失控防范难题，从机理分析、建模预测和安全管理三个角度开展了研究。首先，通过分析动力电池热失控特性，揭示了电池热失控机理，建立了电池热失控特性预测模型。进一步地，揭示了复杂车载工况下电池全生命周期热失控特性演变机制，并对负极析锂引起的电池热失控特性演变规律进行了定量分析与建模。最后，针对负极析锂副反应，建立了电池析锂电化学机理模型，提出了析锂在线检测方法，实现了析锂电池的准确检测。 首先，对动力电池热失控机理开展了研究，通过绝热热失控实验与液氮冷却终止热失控实验，分析了绝热热失控过程中大规模内短路与热失控的关系，揭示了由电池内阻剧增导致的大规模内短路不触发热失控的机理。进一步地，通过对电池组份材料进行差示扫描量热测试，明确了电池热失控过程的主要产热反应，发现了负极材料与电解液产热反应直接触发电池热失控的新机理，并设计卡片电池热失控实验进行了验证。基于组份材料测试结果，对电池热失控主要产热反应进行了动力学分析与建模，建立了电池热失控特性预测模型，实现了基于组份材料产热特性分析的单体电池热失控特性准确预测。 其次，针对电池全生命周期安全性演变，开展了多温度多工况的电池加速寿命试验，并对不同衰减路径下的老化电池进行了热失控测试。结合衰减机理分析与组份材料产热特性测试，揭示了电池全生命周期热失控特性演变机制，指出负极析锂是导致老化电池绝热热失控性能急剧变差的主要原因。进一步地，对析锂引起的电池热失控特性变化机制进行了定量分析，开发了负极析锂量定量表征方法，建立了金属锂与电解液产热反应的动力学模型，准确预测了析锂电池热失控特性的演变。 最后，针对析锂这一严重危害电池安全性的副反应，分析了负极表面锂析出与重嵌入机制，建立了考虑锂析出与重嵌入过程的电池电化学机理模型，开发了基于充电后弛豫电压微分分析的电池充电析锂定量检测方法，实现了电池充电析锂检测，并在实际电池管理系统中进行了验证。进一步开发了基于加热阻抗分析的析锂电池快速筛选方法，实现了析锂电池的快速准确分拣。

With the large-scale application of electric vehicles, safety accidents associated with battery thermal runaway happened occasionally, hindering the development of electric vehicles. The safety performance of lithium-ion power battery system has become not only the core competitiveness of electric vehicles, but also the critical factor that determines the future development of new energy vehicles. Therefore, solutions to mitigate the thermal runaway of power battery system are urgently needed. Thermal runaway of lithium-ion power battery during the whole life cycle is investigated in this Ph.D. Dissertation, focusing on thermal runaway mechanism, model-based prediction of thermal runaway and safety management. The thermal runaway mechanism of lithium-ion battery under adiabatic tests is first revealed. Accurate model-based thermal runaway prediction of lithium-ion battery is achieved based on kinetics analysis of the exothermic reactions between cell components. A comparative study of the evolution of battery thermal runaway behaviors under different degradation paths is conducted. The effects of lithium plating on battery thermal runaway performance are further investigated quantitatively. Furthermore, an electrochemical model of lithium-ion battery incorporated with lithium plating-stripping reactions is established to study the lithium plating mechanism on anode surface. Finally, non-destructive detection methods for lithium plating are proposed based on modeling analysis. Firstly, the thermal runaway mechanism of a 24Ah pouch lithium-ion power battery is investigated using extended volume accelerating rate calorimetry and differential scanning calorimetry (DSC). Massive internal short circuit is found to contribute little to the total heat generation due to the rapid increase of battery resistance during adiabatic thermal runaway process, and thus cannot trigger thermal runaway. Six exothermic reactions in anode-electrolyte and cathode-anode thermodynamic systems turn out to be the dominant heat sources. Particularly, the exothermic reactions between anode and electrolyte are determined as the initiation of battery thermal runaway based on tests of partial cells. Furthermore, kinetics parameters of the six dominant exothermic reactions are identified from DSC tests on cell components. A predictive battery thermal runaway model is then established by superimposing the chemical kinetics models of the six exothermic reactions. The model can accurately predict the adiabatic thermal runaway test results and oven test results of the 24 Ah lithium-ion battery, presenting its capability in determining battery safety performance and optimizing battery chemistry design without producing test batches of batteries. Secondly, a comparative investigation of the aging effects on battery thermal runaway behaviors is implemented. A series of thermal runaway tests on the batteries aged under four different degradation paths are conducted. The correlations between capacity fading mechanism and changes of battery thermal runaway behaviors are revealed from post-mortem analysis on the aged electrodes. Lithium plating is found to be the key reason for the deterioration of battery thermal runaway performance after degradation. In-depth characterization of the effects of lithium plating on battery thermal runaway is carried out. The amount of plated lithium on anode is quantified using 7Li magic-angle spinning nuclear magnetic resonance. Kinetics models of the reactions between plated lithium and electrolyte are built to predict the thermal runaway behaviors of lithium-ion batteries with plated lithium, exhibiting accurate predictions during the whole life cycle. Finally, the lithium plating-stripping behaviors on anode surface and possible detection methods are studied. The lithium plating and stripping reactions are considered as side reactions on the anode surface determined by the anode over-potential. An electrochemical model incorporated with those two reactions is established to investigate the lithium plating-stripping process at low temperature. After validation, the model can successfully predict the characteristic voltage plateau during the rest period after low-temperature charging. Based on modeling analysis, a non-destructive detection method for lithium plating via differential analysis on the voltage plateau is developed. The local minimum occurs in the differential voltage curve is applied as a quantitative indicator for lithium plating. The lithium plating detection method is further verified through tests on a battery module using data collected by a commercial battery management system. Furthermore, a novel method for rapid screening of batteries with plated lithium is proposed by comparing the change of battery impedance before and after short-time heating, achieving excellent performance.