锂离子电池健康状态评估及剩余使用寿命预测技术研究

发布时间：2018-02-02 08:58

本文关键词： 锂离子电池健康状态双卡尔曼滤波模糊推理系统-自适应双卡尔曼滤波剩余使用寿命曲线拟合灰色模型　出处：《南京航空航天大学》2016年硕士论文　论文类型：学位论文

【摘要】：锂离子电池无论是在军用还是民用领域都得到了广泛的应用,准确对其进行健康状态(State of Health,SOH)评估及剩余使用寿命(Remaning Useful Life,RUL)预测对于提高电池安全性与使用寿命具有重要意义。本文以锂离子电池为研究对象,重点开展锂离子电池SOH评估以及RUL预测方法的研究,具体研究内容包括:1)介绍了锂离子电池的工作原理,阐明了锂离子电池常用性能参数的基本概念,概述了锂离子电池的充电方式并通过实验研究了环境温度、放电电流对电池端电压的影响以及电池容量退化规律。2)针对目前卡尔曼滤波算法大多是离线估计电池SOH,无法满足实际工程需要,本文研究了一种基于双卡尔曼滤波算法的电池SOH在线估计方法。首先,通过最小二乘法对电池模型参数进行辨识,实现对电池的建模。然后交替使用两个卡尔曼滤波器分别估计电池的荷电状态与欧姆内阻。为了进一步提高估计精度,本文提出了一种基于模糊推理系统-自适应双卡尔曼滤波(fuzzy inference system-adaptive dual extended Kalman filter,FIS-ADEKF)方法,分别运用Sage-Husa自适应算法与模糊控制器对状态噪声协方差与观测噪声协方差进行修正。最后设计了动态应力测试工况实验进行验证,实验结果表明,改进后的双卡尔曼滤波算法能够实现电池SOH的在线估计,且不依赖于初始值,也不需要事先计算模型参数,具有更高的准确性、收敛性和可行性。3)针对锂离子电池剩余使用寿命预测问题,本文分别利用曲线拟合法与灰色模型来预测电池RUL。在对预测模型进行验证时,本文首先用美国航空航天局艾姆斯研究中心的锂离子电池实验数据来验证,然后通过自主搭建的实验平台获得的实验数据来进一步验证。实验结果表明,一次函数模型与灰色模型能够很好的预测电池RUL,且随着训练数据的增加,预测精度更高。
[Abstract]:Lithium ion batteries have been widely used in both military and civil fields. Soh) evaluation and residual service life / emanation Useful Life. Rul) prediction is of great significance to improve battery safety and service life. This paper focuses on the study of lithium ion battery SOH evaluation and RUL prediction methods. The working principle of lithium ion battery is introduced, and the basic concept of common performance parameters of lithium ion battery is expounded. The charging method of lithium ion battery is summarized and the ambient temperature is studied by experiment. The effect of discharge current on the terminal voltage of the battery and the degradation rule of battery capacity. (2) aiming at the current Kalman filtering algorithm, most of the current Kalman filtering algorithms estimate the SOH of the battery off-line, which can not meet the practical engineering needs. In this paper, a battery SOH online estimation method based on double Kalman filter algorithm is studied. Firstly, the parameters of battery model are identified by least square method. Then, two Kalman filters are used alternately to estimate the charged state and ohmic resistance of the battery. In order to further improve the estimation accuracy. In this paper, an adaptive double Kalman filter based on fuzzy inference system is proposed. Fuzzy inference system-adaptive dual extended Kalman filter. FIS-ADEKF method. Sage-Husa adaptive algorithm and fuzzy controller are used to modify the state noise covariance and observation noise covariance respectively. Finally, the dynamic stress test condition experiment is designed to verify the proposed method. The experimental results show that the improved double Kalman filter algorithm can realize the on-line estimation of battery SOH, and it does not depend on the initial value, nor does it need to calculate the model parameters in advance, so it has higher accuracy. Convergence and Feasibility. 3) aiming at the residual service life prediction of lithium ion battery, this paper uses curve fitting method and grey model to predict the battery RUL.When the prediction model is verified. In this paper, the experimental data of lithium-ion battery from Ames Research Center of NASA are first used to verify, and then the experimental data obtained from the self-built experimental platform are further verified. The experimental results show that. The primary function model and grey model can predict the battery RUL well, and with the increase of training data, the prediction accuracy is higher.
【学位授予单位】：南京航空航天大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TM912

【参考文献】