Application of artificial neural network for the critical flow prediction of discharge nozzle

System thermal-hydraulic (STH) code is adopted for nuclear safety analysis. The critical flow model (CFM) significant the accuracy of STH simulation. To overcome defects current CFMs (low precision or long calculation time), a CFM based on genetic neural network (GNN) has been developed in this work. build powerful model, besides mass flux, pressure and quality were also considered which was seldom before. Comparing with traditional homogeneous equilibrium (HEM) Moody GNN can predict flux higher (approximately 80% results are within ±20% error limit); comparing Leung Shannak prediction, achieved best (more than prediction limit). For quality, similar achieved. GNN-based work meaningful development.