Scholarlink

<p>Nanofluids provide notable benefits in a range of real-world applications because of their improved thermal conductivity. They are used in engine and vehicle systems to increase fuel efficiency and temperature regulation, as well as in electronics cooling to control heat in small, solar thermal collectors, and high-performance devices. They also have promising energy-efficient HVAC systems and advanced nuclear reactor cooling due to their exceptional heat transmission qualities. Additionally, biomedical uses targeted hyperthermia for the treatment of cancer. This research explores various machine learning models, including artificial neural network (ANN), linear regression, decision tree regression, gradient boosting regressor, and XGBoost to forecast Al<sub>2</sub>O<sub>3</sub>–water nanofluid’s thermal conductivity. The ANN model's capacity to represent intricate nonlinear relationships made it worthwhile to incorporate, even though its accuracy was lower than that of XGBoost and gradient boosting. This inclusion strengthens the robustness of the final model selection and contributes to a more comprehensive understanding of how various algorithms perform on the provided dataset. Key parameters such as size of the particle, volume fraction, and temperature are considered as input variables. Additionally, Exploratory Data Analysis and regression graphs are employed to assess the performance of the models. The XGBoost model showed the best performance out of all the other models, achieving a root-mean-square error of 0.016, a mean absolute error of 0.01, and an <i>R</i><sup>2</sup> of 0.991. Furthermore, Shapley additive explanations sensitivity analysis indicated that the most significant factor in forecasting Al<sub>2</sub>O<sub>3</sub>–water nanofluid’s thermal conductivity is particle size.</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

Evaluating machine learning models for predicting thermal conductivity of Al2O3 nanofluids

  • 期刊论文

  • 发表日期：

    2025年7月28日

查看全文
  • Rekha Moolya,
  • P. Rajendra,
  • Shivananda Moolya,
  • Suvarna Kulal

摘要

Nanofluids provide notable benefits in a range of real-world applications because of their improved thermal conductivity. They are used in engine and vehicle systems to increase fuel efficiency and temperature regulation, as well as in electronics cooling to control heat in small, solar thermal collectors, and high-performance devices. They also have promising energy-efficient HVAC systems and advanced nuclear reactor cooling due to their exceptional heat transmission qualities. Additionally, biomedical uses targeted hyperthermia for the treatment of cancer. This research explores various machine learning models, including artificial neural network (ANN), linear regression, decision tree regression, gradient boosting regressor, and XGBoost to forecast Al2O3–water nanofluid’s thermal conductivity. The ANN model's capacity to represent intricate nonlinear relationships made it worthwhile to incorporate, even though its accuracy was lower than that of XGBoost and gradient boosting. This inclusion strengthens the robustness of the final model selection and contributes to a more comprehensive understanding of how various algorithms perform on the provided dataset. Key parameters such as size of the particle, volume fraction, and temperature are considered as input variables. Additionally, Exploratory Data Analysis and regression graphs are employed to assess the performance of the models. The XGBoost model showed the best performance out of all the other models, achieving a root-mean-square error of 0.016, a mean absolute error of 0.01, and an R2 of 0.991. Furthermore, Shapley additive explanations sensitivity analysis indicated that the most significant factor in forecasting Al2O3–water nanofluid’s thermal conductivity is particle size.