Startseite Technik Chapter 12 Data-driven prediction of thermal conductivity ratio in nanoparticle-enhanced 60:40 EG/water nanofluids
Kapitel
Lizenziert
Nicht lizenziert Erfordert eine Authentifizierung

Chapter 12 Data-driven prediction of thermal conductivity ratio in nanoparticle-enhanced 60:40 EG/water nanofluids

  • Himanshu Upreti ORCID logo , Ankita Pandey ORCID logo und Alok Kumar Pandey ORCID logo
Veröffentlichen auch Sie bei De Gruyter Brill
Erkunden Sie dieses Fachgebiet So veröffentlichen Sie bei uns
Flow Dynamics and Heat Transfer
Ein Kapitel aus dem Buch Flow Dynamics and Heat Transfer

Abstract

This study investigates prediction of the thermal conductivity (TC) ratio of three nanofluids,nanofluids i.e., Al2O3 + EG/water, CuO + EG/water, and ZnO + EG/water using various ML (machine learningmachine learning) and NNneural network (neural network) methods. The primary objective is to compare the performance of eleven regression techniques, including RFR (random forest regressor)random forest regressor, LR (linear regression), decision tree regressor, K − neighbor regressor (KNN), support vector regressor, Lasso regression, ridge regression, XG BoostXG boost regressor, artificial neural network, Gaussian Naïve BayesGaussian naïve Bayes, and polynomial regressor, in predicting the TC ratioTC ratio. The dataset, sourced from a thermophysical properties database, consists of 131 rows and 6 columns (corresponding to nanoparticle type, fluid ratio 60:40 EG/water, volume fractionvolume fraction, temperature, nanoparticle size, and TC ratio). The models are evaluated based on performance metrics, for example, MAE (mean absolute error)mean absolute error, MSE (mean square error), and R 2 . The results reveal that the Ridge regression consistently outperforms other models, achieving the highest R 2 score for ZnO and CuO nanofluids, while linear regression shows superior performance for Al2O3 nanofluid. The study demonstrates the potential of MLmachine learning techniques for accurate predictions of nanofluid properties and highlights the importance of model selection in achieving optimal results. The findings provide valuable insights for future applications of MLmachine learning in nanofluid research and engineering. Additionally, an explainable MLmachine learning technique, SHAP, is used to interpret the output obtained from the MLmachine learning models.

Abstract

This study investigates prediction of the thermal conductivity (TC) ratio of three nanofluids,nanofluids i.e., Al2O3 + EG/water, CuO + EG/water, and ZnO + EG/water using various ML (machine learningmachine learning) and NNneural network (neural network) methods. The primary objective is to compare the performance of eleven regression techniques, including RFR (random forest regressor)random forest regressor, LR (linear regression), decision tree regressor, K − neighbor regressor (KNN), support vector regressor, Lasso regression, ridge regression, XG BoostXG boost regressor, artificial neural network, Gaussian Naïve BayesGaussian naïve Bayes, and polynomial regressor, in predicting the TC ratioTC ratio. The dataset, sourced from a thermophysical properties database, consists of 131 rows and 6 columns (corresponding to nanoparticle type, fluid ratio 60:40 EG/water, volume fractionvolume fraction, temperature, nanoparticle size, and TC ratio). The models are evaluated based on performance metrics, for example, MAE (mean absolute error)mean absolute error, MSE (mean square error), and R 2 . The results reveal that the Ridge regression consistently outperforms other models, achieving the highest R 2 score for ZnO and CuO nanofluids, while linear regression shows superior performance for Al2O3 nanofluid. The study demonstrates the potential of MLmachine learning techniques for accurate predictions of nanofluid properties and highlights the importance of model selection in achieving optimal results. The findings provide valuable insights for future applications of MLmachine learning in nanofluid research and engineering. Additionally, an explainable MLmachine learning technique, SHAP, is used to interpret the output obtained from the MLmachine learning models.

Kapitel in diesem Buch

  1. Frontmatter I
  2. Contents V
  3. Aim and scope VII
  4. Preface IX
  5. Acknowledgments
  6. About editors XIII
  7. List of contributing authors XV
  8. Chapter 1 Introduction to flow dynamics and heat transfer 1
  9. Chapter 2 Compressible fluid flow and heat transfer 29
  10. Chapter 3 Non-Newtonian fluid flow and heat transfer 59
  11. Chapter 4 Heat transfer in forced and natural convection 81
  12. Chapter 5 Numerical study of coupled partial differential equations in heat transfer problems with imprecisely defined parameters 91
  13. Chapter 6 Numerical approach to study the effect of uncertain spectrum of field variables in a porous cavity 107
  14. Chapter 7 Investigation of the thermal fluid system using direct numerical simulation 123
  15. Chapter 8 Dynamics of shock-accelerated V-shaped gas interface 139
  16. Chapter 9 Nonlinear and linear analyses of partially ionized plasma 155
  17. Chapter 10 Thermo-fluid behavior of electroosmotic flow in a hydrophobic microchannel under Joule heating and external fields 185
  18. Chapter 11 The study of oscillating water column energy device in a two-layer fluid system of finite impermeable depth 219
  19. Chapter 12 Data-driven prediction of thermal conductivity ratio in nanoparticle-enhanced 60:40 EG/water nanofluids 239
  20. Chapter 13 Industrial applications of flow dynamics and heat transfer 261
  21. Chapter 14 Optimization techniques in flow dynamics and heat transfer 301
  22. Chapter 15 Advanced optimization methods in flow dynamics 335
  23. Index 353
  24. De Gruyter Series in Advanced Mechanical Engineering
https://doi.org/10.1515/9783111661674-012

Kapitel in diesem Buch

  1. Frontmatter I
  2. Contents V
  3. Aim and scope VII
  4. Preface IX
  5. Acknowledgments
  6. About editors XIII
  7. List of contributing authors XV
  8. Chapter 1 Introduction to flow dynamics and heat transfer 1
  9. Chapter 2 Compressible fluid flow and heat transfer 29
  10. Chapter 3 Non-Newtonian fluid flow and heat transfer 59
  11. Chapter 4 Heat transfer in forced and natural convection 81
  12. Chapter 5 Numerical study of coupled partial differential equations in heat transfer problems with imprecisely defined parameters 91
  13. Chapter 6 Numerical approach to study the effect of uncertain spectrum of field variables in a porous cavity 107
  14. Chapter 7 Investigation of the thermal fluid system using direct numerical simulation 123
  15. Chapter 8 Dynamics of shock-accelerated V-shaped gas interface 139
  16. Chapter 9 Nonlinear and linear analyses of partially ionized plasma 155
  17. Chapter 10 Thermo-fluid behavior of electroosmotic flow in a hydrophobic microchannel under Joule heating and external fields 185
  18. Chapter 11 The study of oscillating water column energy device in a two-layer fluid system of finite impermeable depth 219
  19. Chapter 12 Data-driven prediction of thermal conductivity ratio in nanoparticle-enhanced 60:40 EG/water nanofluids 239
  20. Chapter 13 Industrial applications of flow dynamics and heat transfer 261
  21. Chapter 14 Optimization techniques in flow dynamics and heat transfer 301
  22. Chapter 15 Advanced optimization methods in flow dynamics 335
  23. Index 353
  24. De Gruyter Series in Advanced Mechanical Engineering
Jetzt anmelden und 20% sparen
Institutioneller Zugang
Wie funktioniert Authentifizierung?
Haben Sie eine Idee, wie wir unsere Website verbessern können?
Bitte schreiben Sie uns.
© 2025 De Gruyter Brill
Heruntergeladen am 21.11.2025 von https://www.degruyterbrill.com/document/doi/10.1515/9783111661674-012/html?lang=de
Button zum nach oben scrollen