论文部分内容阅读
轧制过程中,针对4200轧机在轧件宽展变化自动预测和控制,分析了轧制过程中宽展变化的影响因素。在神经网络技术和现场实测数据的基础上,利用Matlab人工神经网络工具箱,应用GRNN广义回归神经网络建立宽展变化预测模型来提高轧制宽展变化预测的精度。结果表明,该方法建立的模型可以实现对宽展变化的预测,其预测精度有较大提高。
During the rolling process, automatic prediction and control of 4200 rolling mill rolling width variation was analyzed, and the influencing factors of the wide span variation in the rolling process were analyzed. On the basis of neural network technology and on-site measured data, using the Matlab artificial neural network toolbox, the GRNN generalized regression neural network is used to establish a forecast model of wide-span variation to improve the accuracy of prediction of wide-spread rolling changes. The results show that the model established by this method can realize the prediction of wide-spread variation and the prediction accuracy is greatly improved.