摘要
使用传统的物理化学方法来预测锂电池的健康状态效率低下且精度不高。为此,本文提出使用极限学习机来对蓄电池的健康状态进行预测。首先对提取出的特征数据集进行归一化预处理;然后,在训练集上使用网格搜索技术优化极限学习机的模型参数。在测试集上和其他方法的对比实验结果表明:基于极限学习机的锂电池健康状态预测方法性能优秀,有着实际应用的前景。
The accurate prediction of the state of health（SOH） of Lithium-ion battery is of critical importance for ensuring its safety and reliability.However,using the traditional physical-chemical methods to predict the SOH is extremely time-consuming with low prediction accuracy.In this study,we proposed a new SOH prediction method based on extreme learning machine（ELM）.First,historical battery data are pre-processed and normalized;then,we train an ELM on the training dataset,the parameters of ELM are optimized using a grid-search strategy.Comparison results on the testing dataset demonstrate that the proposed method outperform the existing SOH prediction methods and has a prospect of practical application.
作者
陈婕
金馨
CHEN Jie;JIN Xin(Nanjing No.3 Middle School,Nanjing,Jiangsu,210001)
出处
《软件》
2018年第2期191-196,共6页
Software
关键词
神经网络
极限学习机
锂电池
健康状态
预测
Neural network
Extreme learning machine
Lithium-ion batteries
State of health
Prediction
