Jelatin Çözeltilerinin Dinamik Viskozitesine Yapay Sinir Ağı (YSA) Yaklaşımı: Esnek Hesaplama Çalışması

Öz

Bu araştırmada, bir dizi ölçümden toplanmış deneysel veriyi kullanarak jelatin çözeltilerinin dinamik viskozitesini tahmin etmek üzere geliştirilen çok katmanlı ileri beslemeli bir yapay sinir ağı modeli (YSA) sunuyoruz. YSA yapısında, kayma gerilmesi, kayma oranı, mil torku, mil açısal hızı ile birlikte jelatin çözeltilerinin kütle konsantrasyonu giriş nöronları olarak tanıtılırken, jelatin çözeltilerinin dinamik viskozitesi tahmin edilmek üzere tek bir çıkış nöronu olarak kullanılmıştır. Geliştirilen YSA modeli, Bayesian regülasyonu ile optimize edilmiş geri yayılım algoritması kullanılarak eğitilmiştir. İlk olarak, en doğru performans sonuçlarını veren YSA yapısını bulmak üzere gizli katmanın optimal geometrik yapısı çalışılmıştır. Önerilen ağ modelleri için ortalama karesel hata (MSE), ortalama mutlak hata (MAE), ortalama kare hatalarının karekökü (RMSE), determinasyon katsayısı (R^2), varyans (VAF) ve regresyon analizleri performans değerlendirme parametreleri olarak kullanılmıştır. Geliştirilen YSA modelinin başarı performansını etkileyen en etkin giriş nöronunu araştırmak amacıyla duyarlılık analizi yapılmıştır. Sonuç olarak, gizli katmanda 8 nöronun kullanılması, önerilen diğer YSA modellerine kıyasla en düşük MSE ve en yüksek R^2 değerlerini vererek en yüksek başarı performansını göstermiştir. Duyarlılık analizinin sonucu olarak, kayma oranı oluşturulan sinir ağının başarı performansını etkileyen en etkin giriş nöronu olarak bulunmuştur. Tahmin edilen dinamik viskozite değerlerinin, ölçülen dinamik viskozite değerleriyle büyük bir uyum içinde olması, geliştirilen YSA modelinin doğruluğunu ve güvenilirliğini ispatlamıştır. Bu nedenle geliştirilen YSA modeli, bu araştırmada istatistiksel detayları verilen veri aralığındaki giriş ve çıkış parametrelerini kullanarak, polimer çözeltilerinin dinamik viskozitesini tahmin etmek için efektif bir kullanım sağlamaktadır.

Anahtar Kelimeler

• Yapay zeka,Makine Öğrenmesi,Yapay Sinir Ağları,Bayesian Regülasyonu,Regresyon,Duyarlılık Analizi,İstatistiksel Analiz,Jelatin,Biyopolimer,Reoloji,Dinamik Viskozite

Artificial Neural Network (ANN) Approach for Dynamic Viscosity of Aqueous Gelatin Solutions: A Soft Computing Study

Abstract

In this research, we present a multi-layered feed-forward neural network (ANN) model developed for prediction of dynamic viscosity of aqueous gelatin solutions using experimental data collected from a number of measurements. In ANN architecture, shear stress, shear strain, torque of spindle, the angular velocity of spindle together with mass concentrations of gelatin solutions were introduced as input neurons, whereas dynamic viscosity of aqueous gelatin solutions was assigned as a single output neuron to be predicted. Developed ANN model was trained using backpropagation algorithm optimized with Bayesian regulation. Optimal geometry of the hidden layer was first studied to search out the ANN architecture which yields the most accurate performance results. Mean squared error (MSE), mean absolute error (MAE), root-mean-squared error (RMSE), determination of coefficient (R^2), the variance accounted for (VAF) and regression analyses were used as performance assessment parameters for suggested network models. Sensitivity analysis was carried out to investigate the most effective input neuron strongly influencing the performance of the developed ANN model. As a result, the use of 8 neurons in the hidden layer has shown excellent performance results yielding the least MSE and the highest R^2 values compared to other suggested ANN models. Upon sensitivity analysis, the shear rate was found to be the most effective input neuron significantly affecting network performance. ANN-based predicted dynamic viscosity values were found to be in excellent agreement with measured viscosity values, demonstrating the robustness as well as the accuracy of the developed ANN model. Developed ANN model can, therefore, be effectively used to predict the dynamic viscosity of aqueous polymer solutions using the same input and output parameters in specific data range reported in this paper with statistical details.

Keywords

• Artificial Intelligence,Machine Learning,Artificial Neural Network,Bayesian Regulation,Regression,Relative Importance,Statistical Analysis,Biopolymer,Rheology,Dynamic Viscosity

References

1. Akkoyun, S., Yildiz, N., & Kaya, H. (2019). Neural Network Estimation for Attenuation Coefficients for Gamma-Ray Angular Distribution. Physics of Particles and Nuclei Letters, 16(4), 397-401. doi:10.1134/s1547477119040034
2. Aminian, A. (2017). Predicting the effective viscosity of nanofluids for the augmentation of heat transfer in the process industries. Journal of Molecular Liquids, 229, 300-308. doi: https://doi.org/10.1016/j.molliq.2016.12.071
3. Asteris, P. G., Roussis, P. C., & Douvika, M. G. (2017). Feed-Forward Neural Network Prediction of the Mechanical Properties of Sandcrete Materials. Sensors, 17(6), 1344.
4. Burden, F., & Winkler, D. (2009). Bayesian Regularization of Neural Networks. In D. J. Livingstone (Ed.), Artificial Neural Networks: Methods and Applications (pp. 23-42). Totowa, NJ: Humana Press.
5. Char, C., Padilla, C., Campos, V., Pepczynska, M., Díaz-Calderón, P., & Enrione, J. (2019). Characterization and Testing of a Novel Sprayable Crosslinked Edible Coating Based on Salmon Gelatin. Coatings, 9(10), 595.
6. Demirezen, G., & Fung, A. S. (2019). Application of artificial neural network in the prediction of ambient temperature for a cloud-based smart dual fuel switching system. Energy Procedia, 158, 3070-3075. doi: https://doi.org/10.1016/j.egypro.2019.01.992
7. Erdil, A., & Arcaklioglu, E. (2013). The prediction of meteorological variables using artificial neural network. Neural Computing and Applications, 22(7), 1677-1683. doi:10.1007/s00521-012-1210-0
8. Erzin, Y., & Cetin, T. (2013). The prediction of the critical factor of safety of homogeneous finite slopes using neural networks and multiple regressions. Computers & Geosciences, 51, 305-313. doi: https://doi.org/10.1016/j.cageo.2012.09.003

9. Erzin, Y., & Güneş, N. (2011). The prediction of swell percent and swell pressure by using neural networks. Mathematical and Computational Applications, 16(2), 425-436.
10. Erzin, Y., & Turkoz, D. (2016). Use of neural networks for the prediction of the CBR value of some Aegean sands. Neural Computing and Applications, 27(5), 1415-1426. doi:10.1007/s00521-015-1943-7
11. Fatehi, M.-R., Raeissi, S., & Mowla, D. (2017). Estimation of viscosities of pure ionic liquids using an artificial neural network based on only structural characteristics. Journal of Molecular Liquids, 227, 309-317. doi: https://doi.org/10.1016/j.molliq.2016.11.133
12. Foox, M., & Zilberman, M. (2015). Drug delivery from gelatin-based systems. Expert Opinion on Drug Delivery, 12(9), 1547-1563. doi:10.1517/17425247.2015.1037272
13. Garson, G. D. (1991). Interpreting neural-network connection weights. AI Expert, 6(4), 46-51.
14. Ghatak, A., & Robi, P. S. (2018). Prediction of creep curve of HP40Nb steel using artificial neural network.
15. Neural Computing and Applications, 30(9), 2953-2964. doi:10.1007/s00521-017-2851-9
16. Goh, A. T. C. (1995). Back-propagation neural networks for modeling complex systems. Artificial Intelligence in Engineering, 9(3), 143-151. doi: https://doi.org/10.1016/0954-1810(94)00011-S
17. Gullapalli, R. P. (2010). Soft gelatin capsules (softgels). Journal of Pharmaceutical Sciences, 99(10), 4107-4148. doi:10.1002/jps.22151
18. Haykin, S. (1994). Neural Networks: A Comprehensive Foundation: Prentice Hall PTR.
19. Hemmat Esfe, M., & Abbasian Arani, A. A. (2018). An experimental determination and accurate prediction of dynamic viscosity of MWCNT(%40)-SiO2(%60)/5W50 nano-lubricant. Journal of Molecular Liquids, 259, 227-237.
20. Kayri, M. (2016). Predictive Abilities of Bayesian Regularization and Levenberg–Marquardt Algorithms in Artificial Neural Networks: A Comparative Empirical Study on Social Data. Mathematical and Computational Applications, 21(2), 20.
21. Khalaj, G. (2013). Artificial neural network to predict the effects of coating parameters on layer thickness of chromium carbonitride coating on pre-nitrided steels. Neural Computing and Applications, 23(3), 779-786. doi:10.1007/s00521-012-0994-2
22. Nazari, A., Hajiallahyari, H., Rahimi, A., Khanmohammadi, H., & Amini, M. (2019). Prediction compressive strength of Portland cement-based geopolymers by artificial neural networks. Neural Computing and Applications, 31(2), 733-741. doi:10.1007/s00521-012-1082-3
23. Osorio, F. A., Bilbao, E., Bustos, R., & Alvarez, F. (2007). Effects of Concentration, Bloom Degree, and pH on Gelatin Melting and Gelling Temperatures Using Small Amplitude Oscillatory Rheology. International Journal of Food Properties, 10(4), 841-851. doi:10.1080/10942910601128895
24. Pal, S. K., & Chakraborty, D. (2005). Surface roughness prediction in turning using artificial neural network. Neural Computing & Applications, 14(4), 319-324. doi:10.1007/s00521-005-0468-x
25. Wang, Y., Guo, Z., Qian, Y., Zhang, Z., Lyu, L., Wang, Y., & Ye, F. (2019). Study on the Electrospinning of Gelatin/Pullulan Composite Nanofibers. Polymers, 11(9), 1424.