ANNULAR FLOW USING ARTIFICIAL NEURALNETWORKS

Calculation of heat transfer in fluctuating annular flow is one of the serious challenges for activists inthis field. So far, various methods have been presented to answer this problem, including the use ofartificial intelligence and algorithms based on it. In this research, experimental data related to pipeheat transfer have been used to develop a new model based on a neural network to predict heat transferin fluctuating annular flow. In this study, the prediction of heat transfer from a surface with constantheat flux subjected to fluctuating annular flow was investigated using artificial neural networks. Anexperimental study was conducted to estimate the heat transfer characteristics as a function of someinput parameters, namely frequency, amplitude, heat flux, and filling height. The performance ofneural networks was superior compared to the corresponding power law regressions. Subsequently,artificial neural networks were used to predict data from other researchers, but the results were lessaccurate. In this regard, due to the high capacity of the RBF-type neural network, this structure hasbeen used using real data as input. Finally, to check the efficiency of the neural network model, theresults have been compared with the real sample. Based on the results obtained from the validation ofthe proposed neural network by comparing it with experimental data, the amount of heat transfer hasbeen predicted favorably. The prediction of the increase in heat transfer efficiency with a low errorpercentage (range ۱-۰.۹۹) for regression in comparison with the experimental sample, indicates thesufficient compliance of the proposed model with the real model and the efficiency of the network.This study showed that artificial neural networks can be effectively used to model the heat transfer ofoscillating flow in a vertical annular duct.

کلیدواژه ها:

Heat transfer ، fluctuating annular flow ، functional prediction ، RBF neural network ، Artificial intelligence