-
The prediction of heat transfer coefficients or wall temperatures of heat exchanger tubes is an important research topic in supercritical heat transfer, which is extremely significant for the application of supercritical fluids in industrial production and the design of the entire thermal system. At present, the empirical correlation method is the most widely adopted prediction method, but there is still a significant difference between its predicted value of heat transfer coefficient and the actual data near the pseudo-critical temperature. Therefore, some scholars have proposed using artificial neural networks to predict the heat transfer performance of supercritical fluids in tubes. On the basis of previous researches, this paper further explores the effectiveness of artificial neural network in predicting supercritical heat transfer, focusing on the influence of input parameters on neural network prediction results and the influence of genetic algorithm optimization on the prediction results.
In this research, a neural network prediction model for supercritical R1234ze(E) cooled in horizontal straight tubes is established and compared with the modified D-B heat transfer correlation. The result shows that the input parameter has great impact on the prediction accuracy of BPNN, and not all BPNN input parameter combinations can bring better prediction results than heat transfer correlation. The combination of $\operatorname{Re}_b 、 P r_b 、 \rho_b / \rho_w$、$\bar{C}_p / C_{p w} 、 \lambda_b / \lambda_w 、 \mu_b / \mu_w$ 的预测表现最好,对于试验集的预测结果的AAD和Errormax features the best prediction performance. The AAD and Errormax of the prediction result for the trial set are only 2.02 % and 9.34 %, which are far lower than the prediction deviation of the heat transfer correlation, and the predictions of the trend of h in the high temperature region, the maximum value of h and the position of the peak value of h are more precise than correlation. Moreover, this research compares GA-BP model with BP model under two different fitness value calculation methods to reveal the effectiveness of GA-BP in enhancing the prediction accuracy of supercritical heat transfer. It concludes that, when the same dataset is adopted for network training and fitness value calculation, over-fitting will occur and the GA-BP cannot further improve the prediction accuracy; when different datasets are adopted for network training and fitness value calculation, the generalization ability of the network will be strengthened, the root mean square deviation and the maximum deviation of the prediction result can be further reduced. -
[1] Dittus FW and Boelter LMK. Heat transfer in automobile radiators of the tubular type[J]. 1930 University of California Publications of Engineering 2 443.
[2] Gnielinski V. New equation for heat and mass transfer in turbulent pipe and channel flow[J]. 1976 International Journal of Chemical Engineering 16 359.
[3] Bae Y Y and Kim H Y. Convective heat transfer to CO2 at a supercritical pressure flowing vertically upward in tubes and an annular channel[J]. 2009 Experimental Thermal and Fluid Science 33 329.
[4] Saltanov E. Specifics of forced-convective heat transfer to supercritical CO2 flowing upward in vertical bare tubes[D]. 2015 University of Ontario Institute of Technology.
[5] Kim J K, Jeon H K and Lee J S. Wall temperature measurement and heat transfer correlation of turbulent supercritical carbon dioxide flow in vertical circular/non-circular tubes[J]. 2007 Nuclear Engineering and Design 237 1795.
[6] Kuang G, Ohadi M and Dessiatoun S. Semi-empirical correlation of gas cooling heat transfer of supercritical carbon dioxide in microchannels[J]. 2008 HVACR Research 14 861.
[7] Zhang S J, Xu X X, Liu C, Liu X X, Ru Z P, Dang C B. Experimental and numerical comparison of the heat transfer behaviors and buoyancy effects of supercritical CO2 in various heating tubes[J]. 2020 International Journal of Heat and Mass Transfer 149 119074.
[8] Wang L, Pan Y C, Lee J D, Wang Y, Fu B R, Pan C. Experimental investigation in the local heat transfer of supercritical carbon dioxide in the uniformly heated horizontal miniature tubes[J]. 2020 International Journal of Heat and Mass Transfer 159 120136.
[9] Liu S H, Huang Y P, Liu G X, Wang J F, Leung L K H. Improvement of buoyancy and acceleration parameters for forced and mixed convective heat transfer to supercritical fluids flowing in vertical tubes[J]. 2017 International Journal of Heat and Mass Transfer 106 1144.
[10] Kim D E and Kim M H. Experimental study of the effects of flow acceleration and buoyancy on heat transfer in a supercritical fluid flow in a circular tube[J]. 2010 Nuclear Engineering and Design 240 3336.
[11] Kumar R, Nikam K and Jilte R. A simulation model to predict coal-fired power plant production rate using artificial neural network tool[J]. 2020 Applied Computer Vision and Image Processing 1 150.
[12] Ahmadi M H, Ghazvini M, Maddah H, Kahani M, Pourfarhang S, Pourfarhang A, Heris S Z. Prediction of the pressure drop for CuO/(Ethylene glycol-water) nanofluid flows in the car radiator by means of Artificial Neural Networks analysis integrated with genetic algorithm[J]. 2020 Physica A 546 124008.
[13] Aghayari R, Maddah H, Ahmadi M H, Yan W M, Ghasemi N. Measurement and artificial neural network modeling of electrical conductivity of CuO/glycerol nanofluids at various thermal and concentration conditions[J]. 2018 Energies 11 1190.
[14] Mensah R A, Jiang L, Asante-Okyere S, Xu Q, Jin C. Comparative evaluation of the predictability of neural network methods on the flammability characteristics of extruded polystyrene from microscale combustion calorimetry[J]. 2019 Journal of Thermal Analysis and Calorimetry 138 3055.
[15] Ye K, Zhang Y L, Yang L L, Zhao Y R, Li N, Xi C K. Modeling convective heat transfer of supercritical carbon dioxide using an artificial neural network[J]. 2019 Applied Thermal Engineering 150 686.
[16] Zhu B, Zhu X, Xie J, Xu J L, Liu H. Heat transfer prediction of supercritical carbon dioxide in vertical tube based on artificial neural networks[J]. 2021 Journal of Thermal Sciences 30 1751.
[17] Prasad K S R, Krishna V, Bharadwaj M S, Ponangi B R. Turbulent heat transfer characteristics of supercritical carbon dioxide for a vertically upward flow in a pipe using computational fluid dynamics and artificial neural network[J]. 2022 Journal of Heat Transfer 144 011802.
[18] Ma D, Zhou T, Chen J, Qi S, Shahzad M A, Xiao Z J. Supercritical water heat transfer coefficient prediction analysis based on BP neural network[J]. 2017 Nuclear Engineering and Design 320 400.
[19] Chang W, Chu X, Fareed A F B S, Pandey S, Luo J Y, Weigand B, Laurien E. Heat transfer prediction of supercritical water with artificial neural networks[J]. 2018 Applied Thermal Engineering 131 815.
[20] Sun F, Xie G and Li S. An artificial-neural-network based prediction of heat transfer behaviors for in-tube supercritical CO2 flow[J]. 2021 Applied Soft Computing 102 107110.
[21] Sun F, Xie G, Song J, Li S L, Markides C N. Thermal characteristics of in-tube upward supercritical CO2 flows and a new heat transfer prediction model based on artificial neural networks (ANN)[J]. 2021 Applied Thermal Engineering 194 117067.
[22] Jiang Y R, Hu P* and Ibrahim A. Experimental and numerical investigation on heat transfer characteristics of supercritical R1234ze(E) cooled in horizontal tubes[J]. 2020 International Journal of Heat and Fluid Flow 85 108650.
[23] Dang C B and Hihara E. Predicting the cooling heat transfer coefficient of supercritical CO2 with a small amount of entrained lubricating oil by using the neural network method[J]. 2012 International Journal of Refrigeration 35 1130.
Metrics
- Abstract views: 104
- PDF Downloads: 0
- Cited By: 0