Neural Network Based Diagnosis Method for Looper Height Controller of Hot Strip Mills

In this study, NN based diagnostic system for hot strip mills with looper controller are proposed. Recurrent neural network (RNN) is employed to decide presetting of looper control gains. During elapse of time, deterioration of mechanical characteristics is induced together with that of the control system. To overcome the problem, it is required to diagnose true failure cause and to compensate it. For the purpose, the hierarchical neural network (HNN) is applied. HNN model which enables compensation to the deterioration of mill system can estimate current system parameters such as control gains and mill constants. Through numerical experiments, the effect of the proposed method is ascertained