Implementation of Neural Networks for the Prediction of CHF Location

Prediction of critical heat flux (CHF) location is a challenging issue in power generation industries. A variety of factors influence the occurrence of CHF and different mechanisms are proposed to understand the physical phenomena of CHF. Due to the nonlinear and complex behavior of CHF, these mechanisms cannot include all the factors simultaneously, as a result, some of them remain unresolved. Considering the above issues, this study employs a neural network (NN)-based machine learning method to model and predict CHF location. Becker et al. (An experimental investigation of post dryout heat transfer, 1983) data bank was selected for both training and testing. Hyperparameter tuning is applied to optimize the performance of the model. The model is found to be performing well with the training dataset as well as the testing dataset. The proposed model is helpful for the design engineers and analysts in the power plants for design optimization and to reduce the computational time.