基于模型驱动的锂离子电池剩余寿命预测方法研究

发布时间：2018-07-15 13:56

【摘要】：随着锂离子电池的应用领域越来越广泛,其健康状态的管理成为重要的研究领域。作为一种复杂的电化学系统,锂离子电池的容量会随着充放电周期次数的增加而退化,当其退化锂离子电池的失效阈值时,认为此时锂离子电池的寿命达到终结状态。因此,锂离子电池的剩余寿命(Remaining Useful Life,简称RUL)预测成为健康管理中的重要研究内容,获得了普遍重视。近年来,多种方法应用于锂离子电池剩余寿命预测,以期达到提升RUL预测精度。如何准确的描述容量的退化一直是锂离子电池RUL预测的关键,在基于模型驱动的方法中,常用的模型是循环周期数与容量之间的多项式模型和指数模型,以及后来出现的放电时间与容量之间的双指数模型,但是这些模型都将容量的退化看成是一个平滑的曲线,无法准确的描述出容量退化过程中出现的自充电现象。近年来越来越多模型的建立是基于寻求锂离子电池中的其他健康因子与容量之间的关系,作为其建模的依据,但是至今也没有一个比较适合的模型描述容量的退化。为了解决以上的问题,本文致力于建立一个可以准确描述锂离子电池容量退化的模型,用于预测锂离子电池的RUL。首先,为了研究容量与其他健康因子之间的关系,本文提取了阻抗、温度变化速率以及温度差,在分析了这三个健康因子与容量之间的线性关系之后,选择了与容量线性关系比较强的温度变化速率以及温度差与循环周期数一起建立了两个二元回归模型,用于描述容量的退化。其次,在采用SVM方法对健康因子进行预测的基础上提出了基于两个模型的锂离子电池RUL预测方法。最后在两个模型的RUL预测方法上提出了基于聚类的优化方法以及基于集成学习的优化方法和基于加权线性回归的优化方法。
[Abstract]:With the increasing application of lithium ion batteries, the management of their health status has become an important research field. As a complex electrochemical system, the capacity of lithium-ion battery will degrade with the increase of charging and discharging cycle times. When the failure threshold of the degraded lithium ion battery is considered, it is considered that the lifetime of lithium ion battery reaches the end state. Therefore, the prediction of residual life (RUL) of lithium ion batteries has become an important research content in health management, and has been paid more and more attention. In recent years, many methods have been applied to predict the residual life of lithium ion batteries in order to improve the prediction accuracy of RUL. How to accurately describe the degradation of capacity is always the key to predict the Rul of Li-ion batteries. In model-driven methods, the polynomial model and exponential model between cycle number and capacity are commonly used in model-based methods. But these models regard the degradation of capacity as a smooth curve and can not accurately describe the phenomenon of self-charging in the process of capacity degradation. In recent years, more and more models have been established based on seeking the relationship between other health factors and capacity in lithium-ion batteries as the basis for their modeling, but there is no more suitable model to describe the degradation of capacity. In order to solve the above problems, this paper aims to establish a model that can accurately describe the degradation of lithium-ion battery capacity, which can be used to predict the RULL of lithium-ion battery. First of all, in order to study the relationship between capacity and other health factors, the impedance, temperature change rate and temperature difference are extracted, and the linear relationship between these three health factors and capacity is analyzed. Two binary regression models are established to describe the degradation of capacity by selecting the temperature change rate which has a strong linear relationship with capacity and the temperature difference and cycle number. Secondly, based on the prediction of health factors by SVM, a Rul prediction method based on two models for lithium ion batteries is proposed. Finally, the clustering optimization method, the ensemble learning optimization method and the weighted linear regression optimization method are proposed based on the RUL prediction methods of the two models.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM912

【相似文献】