中南大学学报（自然科学版）2025，Vol.56Issue(5)：2089-2098,10.DOI:10.11817/j.issn.1672-7207.2025.05.033
基于注意力机制的多健康因子锂离子电池容量预测方法
Capacity prediction method of multi-health factor lithium-ion battery based on attention mechanism
摘要
Abstract
With the increase the amount of input data,it is difficult for traditional neural network prediction methods to accurately predict the capacity of lithium-ion batteries.In this paper,a multi-health factor capacity prediction method based on attention mechanism was proposed.Firstly,a convolutional neural network(CNN)-short term memory neural network(LSTM)model was established.Then the attention mechanism was added to the CNN-LSTM model,and the CNN-LSTM-Attention combined prediction model was established,which can simultaneously extract 5 health factors strongly related to battery capacity to achieve accurate capacity prediction of battery.Finally,simulation verification was carried out based on the batteriy aging dataset disclosed by NASA.The simulation results show that compared with the tradition mathods,the proposed method has reduced the mean absolute percentage error and root mean square error to varying degrees,which shows that the proposed method has higher prediction accuracy and better robustness than the traditional method.关键词
注意力机制/健康因子/电池容量预测/神经网络/锂离子电池Key words
attention mechanism/health factor/battery capacity prediction/neural network,lithium-ion battery分类
信息技术与安全科学引用本文复制引用
夏永凯,夏向阳,夏天,陶新明,谭建国,桂勇华,冉成科,周雨风,向文乐..基于注意力机制的多健康因子锂离子电池容量预测方法[J].中南大学学报（自然科学版）,2025,56(5):2089-2098,10.基金项目
国家自然科学基金资助项目(51977014)(Project(51977014)supported by the National Natural Science Foundation of China) （51977014）
