Scholarlink

In battery management systems, accurate prediction of the remaining useful life (RUL) and the state of health (SOH) of the battery is crucial for enhancing battery longevity. In this study, a novel prediction model combining bidirectional LSTM and Transformer is proposed for improving SOH prediction accuracy. First, to enhance the model’s generalization ability and address the issue of gradient vanishing, the bidirectional LSTM algorithm was employed. Then to better utilize the limited data, the Transformer module was added to improve the feature extraction capability. Finally the University of Maryland (CALCE) lithium-ion battery dataset was used to validate this algorithm. In the battery life test, the Bi-LSTM-Transformer model has a residual error of 99.32% and the model training and testing time is less than that of RNN, LSTM, and Bi-LSTM, which provides an accurate prediction model for the state of health (SOH) of the battery as well as the remaining life prediction (RUL).

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Lithium-Ion Battery Health State Assessment Based on Bi-LSTM-Transformer Algorithm

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Chong Li,
  • Hui Dai,
  • Jiaolong Ye,
  • Dianan Liu,
  • Weijie Lin,
  • Guanhao Du,
  • Zhile Yang

摘要

In battery management systems, accurate prediction of the remaining useful life (RUL) and the state of health (SOH) of the battery is crucial for enhancing battery longevity. In this study, a novel prediction model combining bidirectional LSTM and Transformer is proposed for improving SOH prediction accuracy. First, to enhance the model’s generalization ability and address the issue of gradient vanishing, the bidirectional LSTM algorithm was employed. Then to better utilize the limited data, the Transformer module was added to improve the feature extraction capability. Finally the University of Maryland (CALCE) lithium-ion battery dataset was used to validate this algorithm. In the battery life test, the Bi-LSTM-Transformer model has a residual error of 99.32% and the model training and testing time is less than that of RNN, LSTM, and Bi-LSTM, which provides an accurate prediction model for the state of health (SOH) of the battery as well as the remaining life prediction (RUL).