基于GRNN神经网络的中厚板轧机宽展预测模型  

Prediction Model of Width Variation of Heavy and Medium Plate Mill Based on Generalized Regression Neural Network

作　　者：卢慧敏[1] 李大磊[1] 覃寿同[1] 孟令启[1] 

机构地区：[1]郑州大学机械工程学院,郑州450052

出　　处：《机械制造》2007年第3期20-23,共4页Machinery

基　　金：国家自然科学基金资助项目(编号:50175031)

摘　　要：为了实现中厚板轧机在轧制过程中的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,预测精度有较大提高。To automatically predict and control the width variation of the heavy and medium plate mill during milling process, the factors affecting the width variation are analyzed. The Matlab neural network box is used to establish the prediction model of width variation with the aid of GRNN so as to improve the accuracy of width variation prediction. The result has indicated that the prediction of width variation can be implemented with the help of the model established in this method and the accuracy of prediction is improved considerably.

关 键 词：中厚板轧机 宽展 GRNN神经网络 预测 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]