Remaining Useful Life Prediction of PEMFC Based on Error Weighting and Stacked Ensemble

ZHANG Chu; TAO Zihan; LI Xi; WANG Zheng; PENG Tian

doi:10.13334/j.0258-8013.pcsee.241192

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

Remaining Useful Life Prediction of PEMFC Based on Error Weighting and Stacked Ensemble

更新时间：2026-10-12

- Remaining Useful Life Prediction of PEMFC Based on Error Weighting and Stacked Ensemble
- Vol. 45, Issue 20, Pages: 8102-8115(2025)
- 作者机构：
  
  淮阴工学院自动化学院,江苏省,淮安市,223003
- 作者简介：
- 基金信息：
- DOI：10.13334/j.0258-8013.pcsee.241192
  CLC：
- Published：2025
- 稿件说明：
移动端阅览
ZHANG Chu, TAO Zihan, LI Xi, et al. Remaining Useful Life Prediction of PEMFC Based on Error Weighting and Stacked Ensemble[J]. 2025, 45(20): 8102-8115.
DOI：

ZHANG Chu, TAO Zihan, LI Xi, et al. Remaining Useful Life Prediction of PEMFC Based on Error Weighting and Stacked Ensemble[J]. 2025, 45(20): 8102-8115. DOI： 10.13334/j.0258-8013.pcsee.241192.

摘要

为提升质子交换膜燃料电池(proton exchange membrane fuel cells，PEMFC)剩余使用寿命(remaining useful life，RUL)预测的准确性与鲁棒性，该文提出一种基于误差加权的堆叠集成深度预测模型。首先，采用Savitzky-Golay (SG)滤波和eXtreme Gradient Boosting (XGBoost)特征选择技术对数据进行预处理，以优化数据质量。接着，构建一个融合深度置信网络(deep belief network，DBN)、门控循环单元(gated recurrent unit，GRU)和时间卷积网络(temporal convolutional network，TCN)的集成模型，该方法根据每个子模型的预测误差分配权重，并使用随机森林(random forest，RF)模型对加权结果做进一步预测，以求得最优预测效果。通过实验对比分析证明，该集成模型在预测PEMFC未来的退化趋势及剩余寿命方面展现出色的性能，预测结果更加精确。

Abstract

To enhance the accuracy and robustness of estimating the remaining useful life (RUL) of proton exchange membrane fuel cells (PEMFC)

an error-weighted stacked ensemble model is proposed in this paper. The data quality is improved by preprocessing with Savitzky-Golay (SG) filter and feature selection using eXtreme Gradient Boosting (XGBoost). The model integrates the deep belief network (DBN)

the gated recurrent unit (GRU)

and the temporal convolutional network (TCN). Prediction weights are adjusted based on the errors of each constituent model and combined by using random forest (RF) to achieve optimal results. Comparative experimental analysis shows that this integrated approach significantly improves the prediction accuracy of the degradation trend and RUL of PEMFC.

关键词

Keywords

references

Views

683

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

High-voltage and High-power Hybrid Commutated Converter Based on Reverse Blocking Integrated Gate Commutated Thyristor

Surface Water Film Simulation Research on Biomimetic Stator Imitating Diamond Cell Structure

The impact of negative media coverage on innovation in high-tech firms

Federated Split Reinforcement Learning-driven Privacy-preserving Coordinated Operation Method for Distribution Network and Buildings

The impact of green transformational leadership on radical green innovation

Related Author

No data

Related Institution

No data

AI问答

⁰