System identification using dynamic neural networks and its application to plasticating extruders

Proposes a dynamic neural network (DNN) model and gives the conditions under which the DNN has a unique equilibrium point. The synthesis of a DNN for solving quadratic optimization is given. The resultant approach is applied to the identification of a plastic extruder system. The results show that the model gives results that match those of an actual system. The approach proposed is thus shown to be effective.<>