锂离子电池健康状态估计与剩余寿命预测  

Lithium-ion Battery State-of-Health Estimation and Remaining Useful Life Prediction

作　　者：董汉成[1] 凌明祥[1] 王常虹[1] 李清华[1] 

机构地区：[1]哈尔滨工业大学空间控制与惯性技术研究中心,黑龙江哈尔滨150001

出　　处：《北京理工大学学报》2015年第10期1074-1078,共5页Transactions of Beijing Institute of Technology

基　　金：中央高校基本科研业务费专项资金资助项目(HIT.NSRIF.2014031)

摘　　要：针对锂离子电池健康状态(state-of-health,SOH)估计与剩余有效工作时间(remaining useful life,RUL)预测进行探讨.提出了一种利用SOH参数反应电池状况,并且建模预测电池RUL的方法.改进了现有研究成果在RUL预测中不能更新其概率密度的缺陷.同时应用支持向量回归机(SVR-PF)改进标准粒子滤波算法具有粒子贫化效应的缺点.仿真结果表明提出的参数准确地反应了电池的状况,同时也准确地预测了电池的RUL;SVR-PF具有比粒子滤波更强的平滑与预测能力.Lithium-ion batteries are important power sources of the electronic devices. A method based on the SOH （state-of-health） parameters was proposed to estimate the lithium-ion battery SOH and RUL （remaining useful life）. The improvement was made towards the defect that the current research works do not update the probability density during the RUL prediction process. Moreover, the SVR-PF （support vector regression-particle filter） algorithm was applied to improve the degeneracy phenomenon of the standard particle filter. The simulation results show that the proposed parameters estimate the battery SOH well, and accurately predict the RUL; the SVR-PF has good smoothing and prediction capability.

关 键 词：锂离子电池 健康状态 剩余有效工作时间 健康状态变量 支持向量回归机粒子滤波 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]