几种方法在锂电池RUL预测中的对比研究  

Comparative study on different methods in RUL prediction of lithium batteries

作　　者：梁新成[1] 宋胜 张勉 黄国钧 LIANG Xincheng;SONG Sheng;ZHANG Mian;HUANG Guojun(College of Engineering and Technology,Southwest University,Chongqing 400715,China;College of Artificial Intelligence,Southwest University,Chongqing 400715,China)

机构地区：[1]西南大学工程技术学院,重庆400715 [2]西南大学人工智能学院,重庆400715

出　　处：《电源技术》2022年第6期643-646,共4页Chinese Journal of Power Sources

基　　金：重庆市技术创新与应用发展专项(cstc2019jscx-zdztzxX0042);重庆市自然科学基金项目(cstc2021jcyj-msxmX1062)。

摘　　要：通过引入锂电池剩余寿命的描述,获得基于剩余容量的表征方式。为验证结论的可靠性,以NASA PCoE实验中心的锂离子电池实验数据为对象,分别使用二次函数、灰色理论和扩展卡尔曼滤波方法进行了相关的数据处理。结果表明,训练数据量的大小与预测精度相关,且在同等条件下,扩展卡尔曼滤波算法的预测结果最可靠。为了获取更好的剩余寿命(RUL)预测精度,建议未来采用无迹卡尔曼滤波算法(UKF)结合其他算法的联合估计方法。By introducing RUL of lithium-ion battery,the representation was obtained based on the remaining capacity.In order to verify the reliability of the conclusion,the test data from PCoE experimental center of NASA were investigated with algorithms including quadratic function,grey theory and extended Kalman filter.The results indicate that the size of the training data is related to the prediction accuracy,and the prediction result of extended Kalman filter is the most reliable under the same conditions.In order to obtain more precise prediction of lithium battery RUL,it is recommended to use the joint estimation method of unscented Kalman filter(UKF)combined with other algorithms in the future.

关 键 词：锂离子电池 剩余寿命 预测 扩展卡尔曼滤波 

分 类 号：TM912[电气工程—电力电子与电力传动]