Using neural networks to predict parameters in the hot working of aluminum alloys

The ability of an artificial neural network model, using a back-propagation learning algorithm, to predict the flow stress, roll force and roll torque obtained during the hot compression and rolling of aluminum alloys, is studied. It is shown that well-trained neural network models provide fast, accurate and consistent results, making them superior to other predictive techniques. (C) 1999 Elsevier Science S.A. All rights reserved.