基于随机过程的工业锂电池退化模型研究与应用

发布时间：2018-09-18 09:07

【摘要】：锂离子电池作为能量密度高、循环寿命长、充电时间短,自放电率低的新型材料在手持终端等民用领域已经得到广泛的应用。虽然在新能源电车、航空航天等领域也得到广泛应用,但不时报出的锂电池安全事故使得锂离子电池的可靠性和安全性成为工业应用中的关键问题。关于锂电池剩余使用寿命预测的研究在PHM国际会议中占据较大比例,锂电池的退化过程研究成为PHM技术的关键部分。本文选择电池容量作为特征量,研究锂电池剩余使用寿命问题,主要完成了如下研究工作:美国NASA PCOE数据库锂电池实验数据分析,汇总国内外针对该数据集的科研成果,研究锂离子电池的退化因子变量,采用数据驱动方法挖掘锂电池特征量的数据统计特征,总结针对该数据集的研究方法;ARIMA算法模型研究,介绍ARIMA模型基本理论与应用方法,并以锂电池容量退化量为特征量,建立ARIMA预测模型,对锂电池的剩余使用寿命进行预测;随机过程模型研究,锂电池老化是包括正负极材料老化、隔膜与电解液老化等众多原因累积所致,微观上是大量粒子相互作用的结果;宏观上表现为动态非线性的退化过程,依据中心极限定理,应用常用随机过程模型带漂移Wiener过程建立锂电池剩余寿命预测模型;最后采用LabVIEW与MATLAB混合编程技术,设计实现工业锂电池PHM系统,实验结果表明ARIMA模型短期预测具有具有较高的精确度和较强的可行性但不适合长期预测也不能给出预测的置信值;带漂移的维纳过程不但能够给出剩余使用寿命的预测值,同时也能描述该预测值的可信程度。
[Abstract]:As a new type of material with high energy density, long cycle life, short charging time and low self-discharge rate, Li-ion battery has been widely used in civil fields such as handheld terminals. Although it has been widely used in the fields of new energy tram, aerospace and so on, the reliability and safety of lithium-ion battery have become the key problem in industrial application due to the safety accident of lithium battery. The research on the prediction of the residual life of lithium battery occupies a large proportion in the international conference of PHM. The degradation process of lithium battery becomes the key part of PHM technology. In this paper, the battery capacity is chosen as the characteristic quantity, and the remaining service life of lithium battery is studied. The main work is as follows: the analysis of lithium battery experimental data in NASA PCOE database of USA, the summary of domestic and foreign scientific research results on this data set, This paper studies the degradation factor variables of lithium ion batteries, uses data-driven method to mine the statistical characteristics of lithium batteries, summarizes the research method of Arima algorithm model for the data set, and introduces the basic theory and application method of ARIMA model. The ARIMA prediction model is established to predict the remaining service life of lithium battery, and the stochastic process model is used to study the aging of lithium battery, which includes positive and negative electrode material aging. The accumulation of the membrane and electrolyte aging is the result of the interaction of a large number of particles microscopically, the macroscopic performance is a dynamic nonlinear degradation process, according to the central limit theorem, The residual life prediction model of lithium battery is established by using common stochastic process model with drift Wiener process, and the PHM system of industrial lithium battery is designed and implemented by LabVIEW and MATLAB mixed programming technology. The experimental results show that the short-term prediction of ARIMA model has high accuracy and feasibility, but it is not suitable for long-term prediction and can not give the predicted confidence value, and the Wiener process with drift can not only give the prediction value of the remaining service life, It can also describe the reliability of the prediction.
【学位授予单位】：中国科学院大学(中国科学院沈阳计算技术研究所)
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM912

【参考文献】