扩展卡尔曼滤波优化的锂离子电池寿命预测
Lithium-ion battery life prediction based on extended Kalman filter
张涌 1张翔 1李习龙 1张伟 2赵奉奎1
作者信息
- 1. 南京林业大学 汽车与交通工程学院,南京 210037
- 2. 江苏省特种设备安全监督检验研究院吴江分院,江苏 苏州 215200
- 折叠
摘要
锂离子电池剩余使用寿命(remaining useful life,RUL)的准确预测有助于检测电池的健康状况并提高电池工作时的安全性,具有重要的研究意义和实践价值.然而,由于锂离子电池衰退的非线性以及电池模型的复杂性等问题,对电池RUL的预测变得困难.使用双指数退化模型结合扩展卡尔曼滤波(extended Kalman filter,EKF)算法来预测锂离子电池的RUL,利用Matlab进行了锂离子电池RUL的预测仿真,并将仿真结果与NASA容量数据进行了比较分析.仿真结果表明,双指数退化模型结合EKF算法对电池RUL的预测结果与实际结果之间的偏差较小,总体平均绝对误差约为10.9%,具有较高的精度.
Abstract
Accurate prediction of the remaining useful life(RUL)of lithium-ion batteries is of great significance in detecting the battery's health conditions and improving the safety of battery operation.However,predicting the RUL of lithium-ion batteries is difficult due to the non-linearity of battery degradation and the complexity of battery models.This paper employs a dual-exponential decay model integrated with the extended Kalman filter(EKF)algorithm to predict the RUL of lithium-ion batteries.Matlab is employed for the simulation of RUL prediction,and the simulation results are compared and analyzed with NASA capacity data.Our simulation results demonstrate the dual-exponential decay model integrated with the extended Kalman filter algorithm has a small deviation between the prediction and the actual RUL,with an overall average absolute error of about 10.9%,indicating a high accuracy.
关键词
锂离子电池/RUL预测/双指数退化模型/扩展卡尔曼滤波Key words
lithium-ion batteries/RUL prediction/dual-exponential decay model/extended Kalman filter引用本文复制引用
出版年
2024
国家科技期刊平台
NSTL
万方数据