Scholarlink

<p>Efficient thermal management is critical for improving the durability and performance of proton exchange membrane (PEM) fuel cells. In this study, predictive models for the thermophysical properties of zinc oxide (ZnO) nanoparticle-enhanced coolants are developed using computational intelligence methods. Specifically, artificial neural networks (ANN) and curve-fitting techniques are employed to model and predict the thermophysical properties of ZnO/water–ethylene glycol (EG) nanofluids for PEM fuel cell applications. The models were developed using experimentally measured data covering nanoparticle volume fractions of 0-0.5 vol% and operating temperatures of 40-70&#xa0;°C. For ANN prediction, a multilayer perceptron trained using the Levenberg–Marquardt algorithm was employed, with the optimal architecture consisting of a single hidden layer with 12 neurons, yielding a regression coefficient of 0.9802 and a mean square error of 0.00034873, with prediction margin of deviations (MoD) within −0.4%-10%. Further, a third-order polynomial curve-fitting model achieved a superior coefficient of determination of 0.9995 and a root mean square error of 0.0009, with MoD limits of −0.2%-0.5%. A systematic comparison with existing models reported in the literature demonstrates that the proposed methodologies significantly outperform conventional predictive approaches in terms of accuracy and robustness. The findings establish the effectiveness of computational intelligence techniques as reliable design tools for optimizing nano coolant formulations and operating conditions, thereby contributing to the development of high-performance and energy-efficient PEM fuel cell thermal management systems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction of ZnO Nanoparticle-Enhanced Coolant Properties for PEM Fuel Cell Applications Using Computational Intelligence Methods

  • 期刊论文

  • 发表日期：

    2026年2月6日

查看全文
  • S. Manikandan

摘要

Efficient thermal management is critical for improving the durability and performance of proton exchange membrane (PEM) fuel cells. In this study, predictive models for the thermophysical properties of zinc oxide (ZnO) nanoparticle-enhanced coolants are developed using computational intelligence methods. Specifically, artificial neural networks (ANN) and curve-fitting techniques are employed to model and predict the thermophysical properties of ZnO/water–ethylene glycol (EG) nanofluids for PEM fuel cell applications. The models were developed using experimentally measured data covering nanoparticle volume fractions of 0-0.5 vol% and operating temperatures of 40-70 °C. For ANN prediction, a multilayer perceptron trained using the Levenberg–Marquardt algorithm was employed, with the optimal architecture consisting of a single hidden layer with 12 neurons, yielding a regression coefficient of 0.9802 and a mean square error of 0.00034873, with prediction margin of deviations (MoD) within −0.4%-10%. Further, a third-order polynomial curve-fitting model achieved a superior coefficient of determination of 0.9995 and a root mean square error of 0.0009, with MoD limits of −0.2%-0.5%. A systematic comparison with existing models reported in the literature demonstrates that the proposed methodologies significantly outperform conventional predictive approaches in terms of accuracy and robustness. The findings establish the effectiveness of computational intelligence techniques as reliable design tools for optimizing nano coolant formulations and operating conditions, thereby contributing to the development of high-performance and energy-efficient PEM fuel cell thermal management systems.