Data-Driven-Based Forecasting of Two-Phase Flow Parameters in Rectangular Channel

In the current nuclear reactor system analysis codes, interfacial area concentration and void fraction are mainly obtained through empirical relations based on different flow regime maps. present research, data-driven method has been proposed, using four machine learning algorithms (lasso regression, support vector random forest regression back propagation neural network) in field of artificial intelligence to predict some important two-phase parameters rectangular channels, evaluate performance models multiple metrics. The algorithm was found have strongest ability learn from experimental data this study. Test results show that prediction errors model for concentrations fractions all less than 20%, which means target forecasted with good accuracy.