Prediction of Lithium Battery Voltage and State of Charge Using Multi-Head Attention BiLSTM Neural Network

Predicting battery states such as the voltage and state of charge (SOC) can help us monitor lithium batteries more efficiently during usage. This study proposed a predictive model for the lithium battery voltage and SOC by combining a second-order RC equivalent circuit model with a multi-head attent...

Full description

Saved in:
Bibliographic Details
Main Authors: Haiwen Xi, Taolin Lv, Jincheng Qin, Mingsheng Ma, Jingying Xie, Shigang Lu, Zhifu Liu
Format: Article
Language:English
Published: MDPI AG 2025-03-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/15/6/3011
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850203839354896384
author Haiwen Xi
Taolin Lv
Jincheng Qin
Mingsheng Ma
Jingying Xie
Shigang Lu
Zhifu Liu
author_facet Haiwen Xi
Taolin Lv
Jincheng Qin
Mingsheng Ma
Jingying Xie
Shigang Lu
Zhifu Liu
author_sort Haiwen Xi
collection DOAJ
description Predicting battery states such as the voltage and state of charge (SOC) can help us monitor lithium batteries more efficiently during usage. This study proposed a predictive model for the lithium battery voltage and SOC by combining a second-order RC equivalent circuit model with a multi-head attention Bidirectional Long Short-Term Memory (MHA-BiLSTM) neural network. The equivalent circuit model simulates long-term charge–discharge cycles in Simulink, providing essential data for model training. The BiLSTM model, enhanced by the multi-head attention mechanism, is used for accurate short-term predictions of the battery voltage and SOC. The experimental results demonstrate that the proposed MHA-BiLSTM model outperforms other models in the prediction accuracy, achieving an R<sup>2</sup> of 0.91, with the lowest RMSE of 0.0567 and MAPE of 0.0095. This hybrid approach effectively captures the dynamic behavior of the battery and reduces predictive errors, making it a promising solution for battery health monitoring and management.
format Article
id doaj-art-839fad8a04594ec58c7dbb0eef26b260
institution OA Journals
issn 2076-3417
language English
publishDate 2025-03-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-839fad8a04594ec58c7dbb0eef26b2602025-08-20T02:11:25ZengMDPI AGApplied Sciences2076-34172025-03-01156301110.3390/app15063011Prediction of Lithium Battery Voltage and State of Charge Using Multi-Head Attention BiLSTM Neural NetworkHaiwen Xi0Taolin Lv1Jincheng Qin2Mingsheng Ma3Jingying Xie4Shigang Lu5Zhifu Liu6School of Microelectronics, Shanghai University, Shanghai 200444, ChinaShanghai Institute of Space Power-Sources, Shanghai 200245, ChinaShanghai Institute of Ceramics, Chinese Academy of Sciences, Shanghai 201899, ChinaShanghai Institute of Ceramics, Chinese Academy of Sciences, Shanghai 201899, ChinaShanghai Institute of Space Power-Sources, Shanghai 200245, ChinaInstitute for Sustainable Energy, Shanghai University, Shanghai 200444, ChinaSchool of Microelectronics, Shanghai University, Shanghai 200444, ChinaPredicting battery states such as the voltage and state of charge (SOC) can help us monitor lithium batteries more efficiently during usage. This study proposed a predictive model for the lithium battery voltage and SOC by combining a second-order RC equivalent circuit model with a multi-head attention Bidirectional Long Short-Term Memory (MHA-BiLSTM) neural network. The equivalent circuit model simulates long-term charge–discharge cycles in Simulink, providing essential data for model training. The BiLSTM model, enhanced by the multi-head attention mechanism, is used for accurate short-term predictions of the battery voltage and SOC. The experimental results demonstrate that the proposed MHA-BiLSTM model outperforms other models in the prediction accuracy, achieving an R<sup>2</sup> of 0.91, with the lowest RMSE of 0.0567 and MAPE of 0.0095. This hybrid approach effectively captures the dynamic behavior of the battery and reduces predictive errors, making it a promising solution for battery health monitoring and management.https://www.mdpi.com/2076-3417/15/6/3011lithium battery predictionequivalent circuit modelmulti-head attentionbi-directional long short-term memory
spellingShingle Haiwen Xi
Taolin Lv
Jincheng Qin
Mingsheng Ma
Jingying Xie
Shigang Lu
Zhifu Liu
Prediction of Lithium Battery Voltage and State of Charge Using Multi-Head Attention BiLSTM Neural Network
Applied Sciences
lithium battery prediction
equivalent circuit model
multi-head attention
bi-directional long short-term memory
title Prediction of Lithium Battery Voltage and State of Charge Using Multi-Head Attention BiLSTM Neural Network
title_full Prediction of Lithium Battery Voltage and State of Charge Using Multi-Head Attention BiLSTM Neural Network
title_fullStr Prediction of Lithium Battery Voltage and State of Charge Using Multi-Head Attention BiLSTM Neural Network
title_full_unstemmed Prediction of Lithium Battery Voltage and State of Charge Using Multi-Head Attention BiLSTM Neural Network
title_short Prediction of Lithium Battery Voltage and State of Charge Using Multi-Head Attention BiLSTM Neural Network
title_sort prediction of lithium battery voltage and state of charge using multi head attention bilstm neural network
topic lithium battery prediction
equivalent circuit model
multi-head attention
bi-directional long short-term memory
url https://www.mdpi.com/2076-3417/15/6/3011
work_keys_str_mv AT haiwenxi predictionoflithiumbatteryvoltageandstateofchargeusingmultiheadattentionbilstmneuralnetwork
AT taolinlv predictionoflithiumbatteryvoltageandstateofchargeusingmultiheadattentionbilstmneuralnetwork
AT jinchengqin predictionoflithiumbatteryvoltageandstateofchargeusingmultiheadattentionbilstmneuralnetwork
AT mingshengma predictionoflithiumbatteryvoltageandstateofchargeusingmultiheadattentionbilstmneuralnetwork
AT jingyingxie predictionoflithiumbatteryvoltageandstateofchargeusingmultiheadattentionbilstmneuralnetwork
AT shiganglu predictionoflithiumbatteryvoltageandstateofchargeusingmultiheadattentionbilstmneuralnetwork
AT zhifuliu predictionoflithiumbatteryvoltageandstateofchargeusingmultiheadattentionbilstmneuralnetwork