基于Elman网络的金属应力状态系数模型

The Model of Influential Coefficient in Stressed State of Medium and Heavy Plate Rolling Mill Based on Elman Neural Network

以4200mm轧机轧制71块钢板的实测数据为基础,利用Matlab神经网络工具箱,建立了轧制变形区的应力状态系数与轧前厚度、轧后厚度及轧辊直径对应关系的Elman神经网络预测模型。结果表明,Elman神经网络模型有较好的预测精度。

Based on the experimental data obtained from 71 steel plates rolled in 4200 rolling mill, this paper established an Elman neural network prediction model for the relation between influential coefficient in stressed state and thickness before rolling, thickness after rolling and diameter of roller by Matlab neural network toolbox. The results indicate that Elman neural network has better accuracy.