Scholarlink

<p>The optimization of renewable energy systems is a critical area of research, particularly in enhancing the performance of solar energy collectors. In this study, the thermal conductivity and efficiency of phase change material solar collectors enhanced with rice husk graphene were analyzed with a machine learning approach. Two different artificial neural network models were developed using the experimentally obtained data set. There are 12 and 10 neurons in the hidden layers of artificial neural network models with multi-layer architecture. Among the study findings are that the thermal conductivity of paraffin wax increases significantly with the addition of graphene particles, the highest values ​​are observed at high temperatures, and the efficiency of solar collectors increases with increasing solar radiation and flow rates. Thermal conductivity and efficiency values ​​obtained from the network models were compared with the experimental results and a high agreement was observed. The coefficient of determination values ​​calculated for both neural networks were obtained above 0.99. Furthermore, the Index of Agreement values also exceeded 0.99, indicating strong predictive capability. Quantitatively, the Mean Squared Error values were notably low, specifically 4.38E-05 for thermal conductivity and 9.66E-06 for efficiency. Most importantly, the models exhibited remarkably small average deviation rates, with the Margin of Deviation found to be -0.26% for thermal conductivity and -0.09% for efficiency, unequivocally demonstrating their very high accuracy and generalization capabilities in predicting these critical performance measures for rice husk graphene-enhanced phase change material solar collectors.</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

Integrating experimental data with machine learning models to predict thermal conductivity and efficiency in rice husk graphene-enhanced PCM solar collectors

  • 期刊论文

  • 发表日期：

    2025年9月3日

查看全文
  • Andaç Batur Çolak

摘要

The optimization of renewable energy systems is a critical area of research, particularly in enhancing the performance of solar energy collectors. In this study, the thermal conductivity and efficiency of phase change material solar collectors enhanced with rice husk graphene were analyzed with a machine learning approach. Two different artificial neural network models were developed using the experimentally obtained data set. There are 12 and 10 neurons in the hidden layers of artificial neural network models with multi-layer architecture. Among the study findings are that the thermal conductivity of paraffin wax increases significantly with the addition of graphene particles, the highest values ​​are observed at high temperatures, and the efficiency of solar collectors increases with increasing solar radiation and flow rates. Thermal conductivity and efficiency values ​​obtained from the network models were compared with the experimental results and a high agreement was observed. The coefficient of determination values ​​calculated for both neural networks were obtained above 0.99. Furthermore, the Index of Agreement values also exceeded 0.99, indicating strong predictive capability. Quantitatively, the Mean Squared Error values were notably low, specifically 4.38E-05 for thermal conductivity and 9.66E-06 for efficiency. Most importantly, the models exhibited remarkably small average deviation rates, with the Margin of Deviation found to be -0.26% for thermal conductivity and -0.09% for efficiency, unequivocally demonstrating their very high accuracy and generalization capabilities in predicting these critical performance measures for rice husk graphene-enhanced phase change material solar collectors.