锂离子电池充电方法及寿命预测研究

发布时间：2018-03-05 18:33

本文选题：锂离子电池　切入点：充电方法　出处：《北京工业大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着经济发展带来的环境和资源问题日益突出，发展新能源电动汽车已成为汽车产业发展的必由之路。作为电动汽车动力源的电池，其性能直接影响电动汽车的性能。充电方法对电池寿命和电动汽车的推广普及有很大影响，科学的充电方法，可以在不影响动力电池寿命的前提下缩短充电时间；同时，对锂离子电池的健康状态和寿命进行精确预测是保障动力电池系统可靠性的必要基础，对动力电池的延寿维护具有重要的意义。本文研究分析了锂离子电池充电方法，包括恒流恒压充电，脉冲充电，多阶段恒流充电和急充电，分析了充电方法的评价指标；以10Ah的磷酸铁锂软包电池为实验对象，比较分析了脉冲充电方法和恒流充电方法对充电效率的影响；并结合锂离子电池的充电效率和温升情况，提出了限时大电流充电的方法；以恒流恒压充电方法为基础，建立了适用于不同恒流倍率充电的统一充电模型，用于预测剩余充电时间。动力电池寿命与充放电制度、使用温度及电池材料等因素密切相关，其影响关系复杂。结合电池健康状态SOH的内阻定义法，本文以电池等效电模型为基础，在电池充放电数据基础上利用递推最小二乘法对电池的内阻进行参数辨识，用于对电池的健康状态进行预测；结合电池健康程度的容量定义法和动力电池失效阈值的概念，基于电池多次循环的容量数据，分别用最小二乘曲线拟合和相关向量机(Relevance Vector Machine, RVM)的方法，，对锂离子电池剩余循环寿命(Remaining Cycle Life, RCL)进行了回归预测。两种方法比较表明，相关向量机具有较强的局部预测能力，在前期数据足够的情况下，预测电池寿命的精度更高。
[Abstract]:With the environmental and resource problems arising from economic development, the development of new energy electric vehicles has become the only way to the development of the automobile industry. Its performance directly affects the performance of electric vehicle. Charging method has great influence on battery life and popularization of electric vehicle. Scientific charging method can shorten charging time without affecting power battery life. It is necessary to predict the health status and lifetime of Li-ion batteries accurately, which is the necessary basis to ensure the reliability of power battery system, and it is of great significance to the maintenance of battery life extension. This paper studies and analyzes the charging methods of lithium ion battery, including constant current constant voltage charging, pulse charging, multistage constant current charging and rapid charging, and analyzes the evaluation indexes of the charging methods. The effects of pulse charging method and constant current charging method on charging efficiency are compared and analyzed, and combined with the charging efficiency and temperature rise of lithium ion battery, the method of limiting high current charging is put forward, which is based on constant current and constant voltage charging method. A unified charging model suitable for different constant current rate charging is established to predict the remaining charging time. The life of power battery is closely related to charge / discharge system, use temperature and battery material and so on, and its influence is complicated. Combined with the definition of internal resistance of SOH in healthy state of battery, this paper bases on the equivalent electric model of battery. Based on the battery charge and discharge data, the recursive least square method is used to identify the internal resistance of the battery, which is used to predict the battery health status, combining the capacity definition method of battery health degree and the concept of power battery failure threshold. Based on the capacity data of multiple cycles of the battery, the residual cycle life of lithium ion battery is predicted by using the least square curve fitting and the correlation vector machine (RVM) method, respectively. The residual cycle life of the lithium ion battery is also predicted by the regression analysis, and the comparison between the two methods shows that the residual cycle life of the lithium ion battery can be predicted by the method of the least square curve fitting and the correlation vector machine (RVM). Correlation vector machine has strong local prediction ability, and the precision of battery life prediction is higher when the data is sufficient.
【学位授予单位】：北京工业大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM912

【参考文献】