What question did this study set out to answer?

The aim is to enhance the efficiency and accuracy of predicting electrical conductivity in porous layers of reversible fuel cells using machine learning techniques.

April 24, 2026Chinese Science Bulletin (Chinese Version)Open Access

Electrical conductivity of porous layers in reversible fuel cells based on transfer learning

View Full Paper

Ask AI

Bookmark

Authors

YRYiquan Rui BWBowen Wang QDQing Du

Discussion

Member takes

Overview

This research demonstrates high-precision predictions of electrical conductivity in porous layers using advanced machine learning methods, indicating improved efficiency in fuel cell technology.

Key Points

The aim is to enhance the efficiency and accuracy of predicting electrical conductivity in porous layers of reversible fuel cells using machine learning techniques.
Random reconstruction to build microscopic geometric models of carbon paper and titanium felt.
Development of four machine learning models using gradient boosting, support vector regression, kernel ridge regression, and deep neural networks.
Bayesian optimization for adaptive tuning of model hyperparameters and application of transfer learning for small sample prediction.
The Bayesian-optimized model predicts carbon paper electrical conductivity with an R-square of 0.9981.
The transfer learning model achieves an R2 of 0.9979 for titanium felt conductivity predictions.
SHAP analysis reveals that porosity non-linearly influences the conductivity of both materials.

Cite This Study

Rui et al. (2026) studied this question.

synapsesocial.com/papers/69eb0899553a5433e34b3765 https://doi.org/10.1360/csb-2026-0141

View Full Paper

Ask AI

Bookmark

April 24, 2026Chinese Science Bulletin (Chinese Version)Open Access

Electrical conductivity of porous layers in reversible fuel cells based on transfer learning

View Full Paper

Ask AI

Bookmark

Authors

YRYiquan Rui BWBowen Wang QDQing Du

Discussion

Member takes

Overview

This research demonstrates high-precision predictions of electrical conductivity in porous layers using advanced machine learning methods, indicating improved efficiency in fuel cell technology.

Key Points

The aim is to enhance the efficiency and accuracy of predicting electrical conductivity in porous layers of reversible fuel cells using machine learning techniques.
Random reconstruction to build microscopic geometric models of carbon paper and titanium felt.
Development of four machine learning models using gradient boosting, support vector regression, kernel ridge regression, and deep neural networks.
Bayesian optimization for adaptive tuning of model hyperparameters and application of transfer learning for small sample prediction.
The Bayesian-optimized model predicts carbon paper electrical conductivity with an R-square of 0.9981.
The transfer learning model achieves an R2 of 0.9979 for titanium felt conductivity predictions.
SHAP analysis reveals that porosity non-linearly influences the conductivity of both materials.

Cite This Study

Rui et al. (2026) studied this question.

synapsesocial.com/papers/69eb0899553a5433e34b3765 https://doi.org/10.1360/csb-2026-0141

View Full Paper

Ask AI

Bookmark