Hybrid neural network model for the evaluation of the free surface of rolled bar

The application of neural networks to predict the free surface in hot shape rolling process is illustrated in this paper. A round-oval pass has been analyzed because of the opportunity to compare the network performances with empirical models available in bibliography. The aim of this research is the development of a more general model by integrating artificial intelligence and FE techniques in a hybrid experimental-FEM neural network. The performances of different NNs configurations were severely tested to provide a comparison among the networks.