A novel adaptive wavelet network and application

It is difficult for traditional network model to overcome the uncertainties of industrial process. The disadvantage was improved by one novel adaptive wavelet network in a way. First a simple adaptive algorithm was given, and the adaptive performance was analyzed. Then a predictive model for hot finishing rolling natural width change was constructed, including two parts, offline wavelet network model and a simple adaptive algorithm online. Finally the novel adaptive wavelet predictive model was used to forecast the slab width change of the finishing rolling based on practical production data. The simulation results show that this novel method is effective and feasible.