Evaluating machine learning techniques for fluid mechanics: Comparative analysis of accuracy and computational efficiency

Abstract

This study focuses on applying machine learning (ML) techniques to fluid mechanics problems. Various ML techniques were used to create a series of case studies, where their accuracy and computational costs were compared, and behavior patterns in different problem types were analyzed. The goal is to evaluate the effectiveness and efficiency of ML techniques in fluid mechanics and to contribute to the field by comparing them with traditional methods. Case studies were also conducted using Computational Fluid Dynamics (CFD), and the results were compared with those from ML techniques in terms of accuracy and computational cost. For Case 1, after optimizing relevant parameters, the Artificial Neural Network (ANN), Random Forest (RF), and Support Vector Machine (SVM) models all achieved an R² value above 0.9. However, in Case 2, only the ANN method surpassed this threshold, likely due to the limited data available. In Case 3, all models except for Linear Regression (LR) demonstrated predictive abilities above the 0.9 threshold after parameter optimization. The LR method was found to have low applicability to fluid mechanics problems, while SVM and ANN methods proved to be particularly effective tools after grid search optimization.

Keywords

• Fluid mechanics
• Solar dryer
• Artifical neural networks
• Converging-diverging nozzle
• Stirrer

Kaynakça

1. [1] Camargo A, Muniz G, Duare B, Molle B. Applications of computational fluid dynamics in irrigation engineering. Rev. Ciência Agronômica 2020; 51(5): doi: 10.5935/1806 6690.20200097.
2. [2] Çengel Y, Cimbala JM. Fluid mechanics: Fundamentals and applications. 3rd ed. New York: McGraw-Hill Education; 2014.
3. [3] Rosenblatt F. The perceptron: A probabilistic model for information storage and organization in the brain. Psychol. Rev. 1958; 65(6): 386–408. doi: 10.1037/h0042519.
4. [4] Minsky M, Papert SA. Perceptrons: An introduction to computational geometry. Cambridge: The MIT Press; 2017.
5. [5] Hinton GE, Sejnowski TJ. Learning and relearning in Boltzmann machines. In: Rumelhart DE, McClelland JL, editors. Parallel distributed processing: Explorations in the microstructure of cognition. 1st ed. Cambridge: MIT Press; 1986. p. 282–317.
6. [6] LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015; 521(7553): 436–444. doi: 10.1038/nature14539.
7. [7] Mehta UB, Kutler P. Computational aerodynamics and artificial intelligence. California 1984.
8. [8] Abd El-Aziz RM. Renewable power source energy consumption by hybrid machine learning model. Alexandria Eng. J. 2022; 61(12): 9447–9455. doi: 10.1016/j.aej.2022.03.019.

9. [9] Sakthi U, Anil Kumar T, Vimala Kumar K, Qamar S, Kumar Sharma G, Azeem A. Power grid based renewable energy analysis by photovoltaic cell machine learning architecture in wind energy hybridization. Sustain. Energy Technol. Assessments 2023; 57: 103243. doi: 10.1016/j.seta.2023.103243.
10. [10] Tirth V, Algahtani A, Alghtani AH, Al-Mughanam T, Irshad K. Sustainable nanomaterial based technologies for renewable energy production and efficient storage based on machine learning techniques. Sustain. Energy Technol. Assessments 2023; 56: 103085. doi: 10.1016/j.seta.2023.103085.
11. [11] Li Y, Yi YK. Optimal shape design using machine learning for wind energy and pressure. J. Build. Eng. 2023; 70: 106337. doi: 10.1016/j.jobe.2023.106337.
12. [12] Li J, Du X, Martins JRRA. Machine learning in aerodynamic shape optimization. Prog. Aerosp. Sci. 2022; 134: 100849. doi: 10.1016/j.paerosci.2022.100849.
13. [13] Zan BW, Ha ZH, Xu CZ, Liu MQ, Wang WZ. High-dimensional aerodynamic data modeling using a machine learning method based on a convolutional neural network. Adv. Aerodyn. 2022; 4(1): 39. doi: 10.1186/s42774-022-00128-8.
14. [14] Peng W, Zhang Y, Laurendeau E, Desmarais MC. Learning aerodynamics with neural network. Sci. Rep. 2022; 12(1): 6779. doi: 10.1038/s41598-022-10737-4.
15. [15] Khan A, Rajendran P, Sidhu JSS, Thanigaiarasu S, Raja V, Al-Mdallal Q. Convolutional neural network modeling and response surface analysis of compressible flow at sonic and supersonic Mach numbers. Alexandria Eng. J. 2023; 65: 997–1029. doi: 10.1016/j.aej.2022.10.006.
16. [16] Esrafilian-Najafabadi M, Haghighat F. Impact of occupancy prediction models on building HVAC control system performance: Application of machine learning techniques. Energy Build. 2022; 257: 111808. doi: 10.1016/j.enbuild.2021.111808.
17. [17] Ulucak O, Koçak E, Bayer O, Beldek U, Yapıcı EO, Aylı E. Developing and implementation of an optimization technique for solar chimney power plant with machine learning. J. Energy Resour. Technol. 2021; 143(5): doi: 10.1115/1.4050049.
18. [18] Aylı E, Koçak E. Supervised learning method for prediction of heat transfer characteristics of nanofluids. J. Mech. Sci. Technol. 2023; 37(5): 2687–2697. doi: 10.1007/s12206-023-0442-5.
19. [19] Aylı E, Koçak E. Prediction of the heat transfer performance of twisted tape inserts by using artificial neural networks. J. Mech. Sci. Technol. 2022; 36(9): 4849–4858. doi: 10.1007/s12206 022-0843-x.
20. [20] Aylı E. Modeling of mixed convection in an enclosure using multiple regression, artificial neural network, and adaptive neuro-fuzzy interface system models. Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci. 2020; 234(15): 3078–3093. doi: 10.1177/0954406220914330.
21. [21] Sim J, Mohan B, Badra J. Co-optimization of piston bowl and injector for light-duty GCI engine using CFD and ML. Fuel 2022; 329: 125455. doi: 10.1016/j.fuel.2022.125455.
22. [22] Aly AM, Clarke J. Wind design of solar panels for resilient and green communities: CFD with machine learning. Sustain. Cities Soc. 2023; 94: 104529. doi: 10.1016/j.scs.2023.104529.
23. [23] Ye Z, Hsu SC. Predicting real-time deformation of structure in fire using machine learning with CFD and FEM. Autom. Constr. 2022; 143: 104574. doi: 10.1016/j.autcon.2022.104574.
24. [24] Jin X, et al. Exploring the influence of nasal vestibule structure on nasal obstruction using CFD and machine learning method. Med. Eng. Phys. 2023; 117: 103988. doi: 10.1016/j.medengphy.2023.103988.
25. [25] Le DK, Guo M, Yoon JY. A hybrid CFD–deep learning methodology to improve the accuracy of cut-off diameter prediction in coarse-grid simulations for cyclone separators. SSRN Electron. J. 2022. doi: 10.2139/ssrn.4266592.
26. [26] Milićević A, et al. Effects of biomass particles size and shape on combustion process in the swirl-stabilized burner reactor: CFD and machine learning approach. Biomass and Bioenergy 2023; 174: 106817. doi: 10.1016/j.biombioe.2023.106817.
27. [27] Upadhyay M, Nagulapati VM, Lim H. Hybrid CFD-neural networks technique to predict circulating fluidized bed reactor riser hydrodynamics. J. Clean. Prod. 2022; 337: 130490. doi: 10.1016/j.jclepro.2022.130490.
28. [28] Nasution MKM, Elveny M, Syah R, Behroyan I, Babanezhad M. Numerical investigation of water forced convection inside a copper metal foam tube: Genetic algorithm (GA) based fuzzy inference system (GAFIS) contribution with CFD modeling. Int. J. Heat Mass Transf. 2022; 182: 122016. doi: 10.1016/j.ijheatmasstransfer.2021.122016.
29. [29] Mittal G, Kikugawa RI. Computational fluid dynamics simulation of a stirred tank reactor. Mater. Today Proc. 2021; 46: 11015–11019. doi: 10.1016/j.matpr.2021.02.102.
30. [30] Jia Z, Xu L, Duan X, Mao ZS, Zhang Q, Yang C. CFD simulation of flow and mixing characteristics in a stirred tank agitated by improved disc turbines. Chinese J. Chem. Eng. 2022; 50: 95–107. doi: 10.1016/j.cjche.2022.05.017.
31. [31] Li J, Deng B, Zhang B, Shen X, Kim CN. CFD simulation of an unbaffled stirred tank reactor driven by a magnetic rod: assessment of turbulence models. Water Sci. Technol. 2015; 72(8): 1308–1318. doi: 10.2166/wst.2015.314.
32. [32] Sarkar S, Singh KK, Suresh Kumar K, Sreekumar G, Shenoy KT. A novel ANN-CFD model for simulating flow in a vortex mixer. Chem. Eng. Sci. 2022; 260: 117819. doi: 10.1016/j.ces.2022.117819.
33. [33] Shoeibi S, Kargarsharifabad H, Mirjalily SAA, Zargarazad M. Performance analysis of finned photovoltaic/thermal solar air dryer with using a compound parabolic concentrator. Appl. Energy 2021; 304: 117778. doi: 10.1016/j.apenergy.2021.117778.
34. [34] Benhamza A, Boubekri A, Atia A, Hadibi T, Arıcı M. Drying uniformity analysis of an indirect solar dryer based on computational fluid dynamics and image processing. Sustain. Energy Technol. Assessments 2021; 47: 101466. doi: 10.1016/j.seta.2021.101466.
35. [35] Sileshi ST, Hassen AA, Adem KD. Simulation of mixed-mode solar dryer with vertical air distribution channel. Heliyon 2022; 8(11): e11898. doi: 10.1016/j.heliyon.2022.e11898.
36. [36] Chavan A, Vitankar V, Shinde N, Thorat B. CFD simulation of solar grain dryer. Dry. Technol. 2021; 39(8): 1101–1113. doi: 10.1080/07373937.2020.1863422.
37. [37] Singh R, Salhan P, Kumar A. CFD modelling and simulation of an indirect forced convection solar dryer. IOP Conf. Ser. Earth Environ. Sci. 2021; 795(1): 012008. doi: 10.1088/1755 1315/795/1/012008.
38. [38] Sudhakar P, Cheralathan M. Thermal performance enhancement of solar air collector using a novel V-groove absorber plate with pin-fins for drying agricultural products: an experimental study. J. Therm. Anal. Calorim. 2020; 140(5): 2397–2408. doi: 10.1007/s10973-019-08952-9.
39. [39] Karim MA, Hawlader MNA. Performance investigation of flat plate, v-corrugated and finned air collectors. Energy 2006; 31(4): 452–470. doi: 10.1016/j.energy.2005.03.007.
40. [40] Khan SA, Ibrahim OM, Aabid A. CFD analysis of compressible flows in a convergent divergent nozzle. Mater. Today Proc. 2021; 46: 2835–2842. doi: 10.1016/j.matpr.2021.03.074.
41. [41] Subramani N, Sangeetha M, Gowtham G, Sai Kumar A. Flow and acoustic characteristics of convergent-divergent nozzle with and without wedges. Mater. Today Proc. 2023; Apr: 1-7. doi: 10.1016/j.matpr.2023.03.608.
42. [42] Mason ML, Putnam LE, Re RJ. Effect of throat contouring on two-dimensional convergingdiverging nozzles at static conditions. 1980.
43. [43] Brunton SL. Applying machine learning to study fluid mechanics. Acta Mech. Sin. 2021;37(12): 1718-1726. doi: 10.1007/s10409-021-01143-6.
44. [44] Todorovski L, Džeroski S. Combining classifiers with meta decision trees. Mach. Learn. 2003; 50(3): 223-249. doi: 10.1023/A:1021709817809.
45. [45] Zheng X, Schweickert R. Differentiating dreaming and waking reports with automatic text analysis and support vector machines. Conscious. Cogn. 2023; 107: 103439. doi: 10.1016/j.concog.2022.103439.
46. [46] Roy A, Chakraborty S. Support vector machine in structural reliability analysis: A review. Reliab. Eng. Syst. Saf. 2023; 233: 109126. doi: 10.1016/j.ress.2023.109126.
47. [47] Ataseven B. Yapay sinir ağları ile öngörü modellemesi. Öneri Derg. 2013; 10(39): 101-115.
48. [48] Kouvaras N, Dhanak MR. Machine learning based prediction of wave breaking over a fringing reef. Ocean Eng. 2018; 147: 181-194. doi: 10.1016/j.oceaneng.2017.10.005.