Property Prediction Base On Artificial Neural Network And Roll Parameters Optimization

The computer simulation of microstructural evolution in hot rolling and the research of the relationship between microstructure and mechanical properties have been one of the hot topics in recent years。Neural network as a new tool has arouse research wave.In this paper, based on the theory of neural network and combined with Q345E steel which made in Wuhan Iron and Steel Company ,found the BP mapping model and the mechanical properties was predicted also. After referring to a large number of national and international articles published by numerous metallurgical workers and other co-workers'research results, combined the accuracy of research result, some rolling parameters which the most determine the mechanical property of the products was selected and the predict net was programmed. The input parameters of the programe was selected as follows: the heating time, the roughing temperature, the finishing start temperature, the finishing end temperature, the coiling temperature, the reduction and the speed of the finishing out stands. The output paremeter of the programe was mechanical properties. Used the trained network to do the research about the influence of the roll parameters to the mechanical properties. The results of the research showed that using the neural network to predict WL510 steel's mechanical properties can get satisfied results. It can be developed into the on-lined control application.