Abstract: The real-time prediction of BHF control law is a key technology in intelligent deep drawing. Non-axisymmetrical workpiece is studied such as rectangular box and as a result the optimized trend of variable blank holder force (BHF) is obtained which is called fracture critical BHF theory. In this work the result of numerical simulation and experiments are the evidence. The optimized control law of variable BHF is predicted by using feed-forward artificial neural network(ANN), which is in good agreement with experimental results. This measure will ensure the intellectualized control for the whole process of deep drawing.

Key words: Artificial neural network, Intelligent deep drawing, Law of blank holder force, Real-time prediction, Rectangular box