一种航天器锂离子电池寿命预测方法  

Life Prediction Method for Spacecraft Lithium-ion Battery

作　　者：朱立颖[1] 乔明[1] 王涛[2] 陈世杰[1] 

机构地区：[1]北京空间飞行器总体设计部,北京100094 [2]上海空间电源研究所,上海200233

出　　处：《航天器工程》2016年第2期46-51,共6页Spacecraft Engineering

基　　金：国家自然科学基金(51407008)

摘　　要：针对锂离子电池的空间应用特性,根据在轨可测的锂离子电池放电终压,建立寿命预测模型(非平稳自回归模型,ARI模型);将加速因子引入ARI模型中,获得加速非平稳自回归(ND-ARI)模型,通过对放电终压的预测实现电池的寿命预测。利用低地球轨道(LEO)航天器锂离子电池长寿命模拟验证试验数据,分别以原始数据的前30%,50%,70%的训练数据作为模型预测的起始点,对预测方法进行验证。结果表明:在不同建模预测起始点下,ND-ARI模型得到的预测结果具有较高的精度,相对误差均小于0.60%。因此,ND-ARI模型可用于航天器锂离子电池的退化状态监测及寿命预测体系的构建。This paper presents a lithium-ion battery life prediction model ARI（auto regressive integrated）model according to measurable discharge ending voltage on orbit which is characteristics of the lithium-ion battery for space applications.The acceleration factor is imported into the ARI model to obtain a ND-ARI model.The ND-ARI model can accurately predict the discharge ending voltage which represents the life of the lithium-ion battery.The test data of long life LEO spacecraft lithium-ion battery are used to verify the ND-ARI model,and the first 30%,50%and 70%of the test data are used as the start points.The results show that the ND-ARI model predictions at different start points of the training data have high precision,and the relative errors are less than0.60%.The ND-ARI model can be used in spacecraft lithium-ion battery degraded condition monitoring and life prediction systems.

关 键 词：航天器锂离子电池 寿命预测 放电终压 加速非平稳自回归模型 

分 类 号：V442[航空宇航科学与技术—飞行器设计]