The Application of a Hybrid Model

In order to improve the prediction precision of flow stress, in view of intrinsic limitation of traditional models, a new method combining the combinative algorithm of group method and modified error function BP networks with mathematical models (Hybrid Model) to predict flow stress is proposed. By simulation, the results show that this method can correctly recur to the flow stress in the sampled data and it can also predict well the non-sampled data. The predicted results with this method are much better than those with the method combining neural networks with mathematical models.