A Data-Driven Approach to Lifespan Prediction for Vehicle Fuel Cell Systems

You are here: Homepage > Search

A Data-Driven Approach to Lifespan Prediction for Vehicle Fuel Cell Systems

Wang, Yupeng / Wang, Kai / Wang, Bowen / Yin, Yan / Zhao, Honghui / Han, Linghai / Jiao, Kui

The durability of proton exchange membrane fuel cell (PEMFC) is a major concern that limits their commercial application. Fuel cells are characterized by a complex internal mechanism and a strong coupling, rendering them susceptible to performance degradation and health issues, which have received increasing attention. However, the degradation of stack performance cannot fully characterize the decline in system performance. This article proposes an aging index based on the dynamic degradation of fuel cell performance under different conditions to predict the performance degradation of PEMFC. Considering the influence of reversible performance degradation and system failure on performance degradation, a degradation prediction method based on a long short-term memory (LSTM) network is proposed. Different operating conditions and experimental datasets validated the performance of the proposed approach. The root-mean-square error (RMSE) for the proposed method is 0.5273 for 2000 h test data, which verifies its accuracy. By matching and optimizing the air compressor and fuel cell operating points, the power and thermal power are used as the prediction limit value to predict the performance of the PEMFC system. It has important guiding significance for the strategic optimization of the fuel cell system and vehicle powertrain.

Access

Check access

Check availability in my library

Order at Subito €

Export, share and cite

Document information

Title :

A Data-Driven Approach to Lifespan Prediction for Vehicle Fuel Cell Systems

Contributors:

Wang, Yupeng (author) / Wang, Kai (author) / Wang, Bowen (author) / Yin, Yan (author) / Zhao, Honghui (author) / Han, Linghai (author) / Jiao, Kui (author)

Published in:

IEEE Transactions on Transportation Electrification ; 9 , 4 ; 5049-5060

Publication date :

2023-12-01

Size :

2182975 byte

ISSN :

2332-7782 , 2372-2088

DOI :

https://doi.org/10.1109/TTE.2023.3237219

Type of media :

Article (Journal)