Degradation Prediction of Proton Exchange Membrane Fuel Cell Based on Multi-Head Attention Neural Network and Transformer Model

Proton exchange membrane fuel cells are a clean energy technology with wide application in transportation and stationary energy systems. Due to the problem of voltage degradation under long-term dynamic loads, predicting their performance degradation trend is of great significance for extending the...

Full description

Saved in:
Bibliographic Details
Main Authors: Yikai Tang, Xing Huang, Yanju Li, Haoran Ma, Kai Zhang, Ke Song
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:Energies
Subjects:
Online Access:https://www.mdpi.com/1996-1073/18/12/3177
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850156346324811776
author Yikai Tang
Xing Huang
Yanju Li
Haoran Ma
Kai Zhang
Ke Song
author_facet Yikai Tang
Xing Huang
Yanju Li
Haoran Ma
Kai Zhang
Ke Song
author_sort Yikai Tang
collection DOAJ
description Proton exchange membrane fuel cells are a clean energy technology with wide application in transportation and stationary energy systems. Due to the problem of voltage degradation under long-term dynamic loads, predicting their performance degradation trend is of great significance for extending the life of proton exchange membrane fuel cells and improving system reliability. This study adopts a data-driven approach to construct a degradation prediction model. In view of the problem of many input parameters and complex distribution of degradation features, a neural network model based on a multi-head attention mechanism and class token is first proposed to analyze the impact of different operating parameters on the output voltage prediction. The importance of each input variable is quantified by the attention weight matrix to assist feature screening. Subsequently, a prediction model is constructed based on Transformer to characterize the voltage degradation trend of fuel cells under dynamic conditions. The experimental results show that the root mean square error and mean absolute error of the model in the test phase are 0.008954 and 0.006590, showing strong prediction performance. Based on the importance evaluation provided by the first model, 11 key parameters were selected as inputs. After this input simplification, the model still maintained a prediction accuracy comparable to that of the full-feature model. This result verifies the effectiveness of the feature screening strategy and demonstrates its contribution to improved generalization and robustness.
format Article
id doaj-art-dd26ca0d2ce14c579b8e5538517e8e0f
institution OA Journals
issn 1996-1073
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series Energies
spelling doaj-art-dd26ca0d2ce14c579b8e5538517e8e0f2025-08-20T02:24:34ZengMDPI AGEnergies1996-10732025-06-011812317710.3390/en18123177Degradation Prediction of Proton Exchange Membrane Fuel Cell Based on Multi-Head Attention Neural Network and Transformer ModelYikai Tang0Xing Huang1Yanju Li2Haoran Ma3Kai Zhang4Ke Song5School of Automotive Studies, Tongji University, Shanghai 201804, ChinaSchool of Automotive Studies, Tongji University, Shanghai 201804, ChinaSchool of Automotive Studies, Tongji University, Shanghai 201804, ChinaSchool of Automotive Studies, Tongji University, Shanghai 201804, ChinaSchool of Aerospace Engineering and Applied Mechanics, Tongji University, Shanghai 200092, ChinaSchool of Automotive Studies, Tongji University, Shanghai 201804, ChinaProton exchange membrane fuel cells are a clean energy technology with wide application in transportation and stationary energy systems. Due to the problem of voltage degradation under long-term dynamic loads, predicting their performance degradation trend is of great significance for extending the life of proton exchange membrane fuel cells and improving system reliability. This study adopts a data-driven approach to construct a degradation prediction model. In view of the problem of many input parameters and complex distribution of degradation features, a neural network model based on a multi-head attention mechanism and class token is first proposed to analyze the impact of different operating parameters on the output voltage prediction. The importance of each input variable is quantified by the attention weight matrix to assist feature screening. Subsequently, a prediction model is constructed based on Transformer to characterize the voltage degradation trend of fuel cells under dynamic conditions. The experimental results show that the root mean square error and mean absolute error of the model in the test phase are 0.008954 and 0.006590, showing strong prediction performance. Based on the importance evaluation provided by the first model, 11 key parameters were selected as inputs. After this input simplification, the model still maintained a prediction accuracy comparable to that of the full-feature model. This result verifies the effectiveness of the feature screening strategy and demonstrates its contribution to improved generalization and robustness.https://www.mdpi.com/1996-1073/18/12/3177proton exchange membrane fuel celldata-driven modeldegradation predictionmulti-head attentiontransformer model
spellingShingle Yikai Tang
Xing Huang
Yanju Li
Haoran Ma
Kai Zhang
Ke Song
Degradation Prediction of Proton Exchange Membrane Fuel Cell Based on Multi-Head Attention Neural Network and Transformer Model
Energies
proton exchange membrane fuel cell
data-driven model
degradation prediction
multi-head attention
transformer model
title Degradation Prediction of Proton Exchange Membrane Fuel Cell Based on Multi-Head Attention Neural Network and Transformer Model
title_full Degradation Prediction of Proton Exchange Membrane Fuel Cell Based on Multi-Head Attention Neural Network and Transformer Model
title_fullStr Degradation Prediction of Proton Exchange Membrane Fuel Cell Based on Multi-Head Attention Neural Network and Transformer Model
title_full_unstemmed Degradation Prediction of Proton Exchange Membrane Fuel Cell Based on Multi-Head Attention Neural Network and Transformer Model
title_short Degradation Prediction of Proton Exchange Membrane Fuel Cell Based on Multi-Head Attention Neural Network and Transformer Model
title_sort degradation prediction of proton exchange membrane fuel cell based on multi head attention neural network and transformer model
topic proton exchange membrane fuel cell
data-driven model
degradation prediction
multi-head attention
transformer model
url https://www.mdpi.com/1996-1073/18/12/3177
work_keys_str_mv AT yikaitang degradationpredictionofprotonexchangemembranefuelcellbasedonmultiheadattentionneuralnetworkandtransformermodel
AT xinghuang degradationpredictionofprotonexchangemembranefuelcellbasedonmultiheadattentionneuralnetworkandtransformermodel
AT yanjuli degradationpredictionofprotonexchangemembranefuelcellbasedonmultiheadattentionneuralnetworkandtransformermodel
AT haoranma degradationpredictionofprotonexchangemembranefuelcellbasedonmultiheadattentionneuralnetworkandtransformermodel
AT kaizhang degradationpredictionofprotonexchangemembranefuelcellbasedonmultiheadattentionneuralnetworkandtransformermodel
AT kesong degradationpredictionofprotonexchangemembranefuelcellbasedonmultiheadattentionneuralnetworkandtransformermodel