新能源汽车是目前行业热点，其技术核心是车用动力电池，而电池快速充电是一大难题。充电的限速步骤不在充电机而在电池自身，单纯增加充电功率而不考虑电池对功率的接受能力将严重损害电池安全和寿命。因此，开发电池安全快充策略是突破充电瓶颈的关键。本文以大容量车用锂离子电池为对象，从机理研究、模型构建与参数辨识、状态观测与优化控制三个方面对充电安全问题进行剖析，最终开发了无析锂快充策略。对充电滥用条件下电池衰减进行了研究和机理辨识，建立了集总参数降维负极电位估计模型，借助开发的稳定内部电位传感器提出了基于电极分解和频域分解的多阶段集总电化学参数辨识方法。基于降维模型设计了负极电位观测器及自适应时变电流安全快充策略，实现了电池无析锂安全快充，并提出了无析锂意义下的时间最优充电策略。 首先，研究了充电滥用下电池衰减机理，阐释了负极析锂反应机制。通过低温加速寿命实验，分析并总结了电池“非线性”衰减规律。基于原位分析和材料形貌表征辨识了充电衰减核心机理为负极析锂导致的活性锂离子损失。针对衰减后的容量恢复现象，基于电压微分和内部电位信号，研究了锂析出后的重嵌入现象，总结了完整的析锂反应机制，明确了抑制析锂的关键在于控制负极电位。第二，建立了集总参数降维负极电位估计模型，开发了集总电化学参数辨识方法。从全维电化学模型出发，通过模型重构得到了电化学模型最小参数集，进而通过推导状态变量的传递函数建立了降维负极电位估计模型。构建了考虑非理想颗粒效应修正的固液相界面通用频域模型，提升了降维模型精度。提高了内部电位传感器稳定性，通过内置式传感器阻隔效应分析，提出了精确测量的方法。依托传感器提出了基于电极分解和频域分解的分阶段参数辨识方法。最后，开发了基于负极电位观测器和电流在线控制器的安全无损快充策略。基于负极电位估计模型，开发了负极电位闭环观测器；基于负极电位在线观测开发了电流在线优化控制器，通过观测器和控制器耦合控制实现了安全快充。针对虚拟电池和商用电池开展了快充测试。耐久性实验表明充电策略安全无析锂。基于安全快充实验结果，提出了锂离子电池的最优充电原则，给出了最优充电曲线的解析表达式，对充电效果进行了验证。
New energy vehicles have gained traction. The key technology involves lithium-ion batteries, whose ability to fast charge is limited. Simply increasing the charging power may sacrifice the safety and lifetime of the battery, since the rate-limiting step is due to the battery performance rather than the charger. Safe and fast charging strategy is in demand for protecting the batteries.In this dissertation, the large-format lithium-ion batteries for vehicle use are focused. The safe and fast charging is achieved step by step from mechanism investigation, model reformulation to state estimation and optimal control. The degradation mechanism of the battery under different charging conditions is identified. A lumped-parameter model for anode potential estimation is built and transfer functions are derived to generate a reduced-order model. A step-wise parameter identification strategy based on electrode and frequency decomposition is proposed with the help of a novel long-term electrode potential sensor. Deposition-free fast charging is successfully realized through the use of a closed-loop anode potential observer and an online current modification controller. Firstly, the mechanism accounting for battery degradation under charge abuse is investigated, and the behavior mechanism of lithium deposition is elucidated. The nonlinear aging phenomenon of the battery is analyzed through the low-temperature accelerated life test. Based on in situ analysis and SEM, the major aging mechanism induced by charging is identified as loss of lithium inventory due to lithium deposition. A deeper look at the capacity recovery phenomenon suggests the lithium reintercalation after plating, which is also investigated by voltage differential and electrode potential signal. A comprehensive lithium deposition reaction mechanism is summarized. Controlling the anode potential is crucial to address lithium deposition.Secondly, a lumped-parameter anode potential estimation model is established, and a novel parameter identification method is developed. Beginning by the full-order P2D model, the minimum parameter set of the P2D model is obtained by model reformulation, and a reduced-order anode potential estimation model is established by deriving the transfer functions of the state variables. A generalized frequency-domain model of solid/electrolyte interface is presented, modified by introducing CPEs to accommodate non-ideal capacitance and particle effects. A long-term electrode potential sensor is invented for parameter identification. The inherent blocking effect of the sensor thoroughly examined to ensure measurement accuracy. A novel identification method based on electrode decomposition and frequency-domain decomposition is proposed with the help of the sensor.Finally, a safe and non-destructive fast charging strategy using the anode potential observer and current modification controller is proposed. The closed-loop anode potential observer is developed based on the reduced-order model. The current modification controller is developed based on the observed anode potential. Safe and fast charging is realized by coupling control. Fast charging tests are executed on both virtual and real batteries. Durability experiments show that the charging strategy is plating-free. Based on the results of the fast charging, an optimal charging theory of lithium-ion batteries is proposed. The analytical expression of the optimal charging current is provided and the validity of the theory is examined.
